Bump Citus version to 13.2devel

13.1.0 https://github.com/citusdata/citus/pull/8006
13.0.4 https://github.com/citusdata/citus/pull/8005
12.1.8 https://github.com/citusdata/citus/pull/8004

.codeclimate.yml

@@ -0,0 +1,7 @@
+exclude_patterns:
+  - "src/backend/distributed/utils/citus_outfuncs.c"
+  - "src/backend/distributed/deparser/ruleutils_*.c"
+  - "src/include/distributed/citus_nodes.h"
+  - "src/backend/distributed/safeclib"
+  - "src/backend/columnar/safeclib"
+  - "**/vendor/"

.devcontainer/.gdbinit

@@ -0,0 +1,33 @@
+# gdbpg.py contains scripts to nicely print the postgres datastructures
+# while in a gdb session. Since the vscode debugger is based on gdb this
+# actually also works when debugging with vscode. Providing nice tools
+# to understand the internal datastructures we are working with.
+source /root/gdbpg.py
+
+# when debugging postgres it is convenient to _always_ have a breakpoint
+# trigger when an error is logged. Because .gdbinit is sourced before gdb
+# is fully attached and has the sources loaded. To make sure the breakpoint
+# is added when the library is loaded we temporary set the breakpoint pending
+# to on. After we have added out breakpoint we revert back to the default
+# configuration for breakpoint pending.
+# The breakpoint is hard to read, but at entry of the function we don't have
+# the level loaded in elevel. Instead we hardcode the location where the
+# level of the current error is stored. Also gdb doesn't understand the
+# ERROR symbol so we hardcode this to the value of ERROR. It is very unlikely
+# this value will ever change in postgres, but if it does we might need to
+# find a way to conditionally load the correct breakpoint.
+set breakpoint pending on
+break elog.c:errfinish if errordata[errordata_stack_depth].elevel == 21
+set breakpoint pending auto
+
+echo \n
+echo ----------------------------------------------------------------------------------\n
+echo when attaching to a postgres backend a breakpoint will be set on elog.c:errfinish \n
+echo it will only break on errors being raised in postgres \n
+echo \n
+echo to disable this breakpoint from vscode run `-exec disable 1` in the debug console \n
+echo this assumes it's the first breakpoint loaded as it is loaded from .gdbinit \n
+echo this can be verified with `-exec info break`, enabling can be done with \n
+echo `-exec enable 1` \n
+echo ----------------------------------------------------------------------------------\n
+echo \n

.devcontainer/.gitignore

@@ -0,0 +1 @@
+postgresql-*.tar.bz2

.devcontainer/.psqlrc

@@ -0,0 +1,7 @@
+\timing on
+\pset linestyle unicode
+\pset border 2
+\setenv PAGER 'pspg --no-mouse -bX --no-commandbar --no-topbar'
+\set HISTSIZE 100000
+\set PROMPT1 '\n%[%033[1m%]%M %n@%/:%> (PID: %p)%R%[%033[0m%]%# '
+\set PROMPT2 '  '

.devcontainer/.vscode/Pipfile

@@ -0,0 +1,12 @@
+[[source]]
+url = "https://pypi.org/simple"
+verify_ssl = true
+name = "pypi"
+
+[packages]
+docopt = "*"
+
+[dev-packages]
+
+[requires]
+python_version = "3.9"

.devcontainer/.vscode/Pipfile.lock

@@ -0,0 +1,28 @@
+{
+    "_meta": {
+        "hash": {
+            "sha256": "6956a6700ead5804aa56bd597c93bb4a13f208d2d49d3b5399365fd240ca0797"
+        },
+        "pipfile-spec": 6,
+        "requires": {
+            "python_version": "3.9"
+        },
+        "sources": [
+            {
+                "name": "pypi",
+                "url": "https://pypi.org/simple",
+                "verify_ssl": true
+            }
+        ]
+    },
+    "default": {
+        "docopt": {
+            "hashes": [
+                "sha256:49b3a825280bd66b3aa83585ef59c4a8c82f2c8a522dbe754a8bc8d08c85c491"
+            ],
+            "index": "pypi",
+            "version": "==0.6.2"
+        }
+    },
+    "develop": {}
+}

.devcontainer/.vscode/generate_c_cpp_properties-json.py

@@ -0,0 +1,84 @@
+#! /usr/bin/env pipenv-shebang
+"""Generate C/C++ properties file for VSCode.
+
+Uses pgenv to iterate postgres versions and generate
+a C/C++ properties file for VSCode containing the
+include paths for the postgres headers.
+
+Usage:
+  generate_c_cpp_properties-json.py <target_path>
+  generate_c_cpp_properties-json.py (-h | --help)
+  generate_c_cpp_properties-json.py --version
+
+Options:
+  -h --help     Show this screen.
+  --version     Show version.
+
+"""
+import json
+import subprocess
+
+from docopt import docopt
+
+
+def main(args):
+    target_path = args['<target_path>']
+
+    output = subprocess.check_output(['pgenv', 'versions'])
+    # typical output is:
+    #      14.8      pgsql-14.8
+    #  *   15.3      pgsql-15.3
+    #      16beta2    pgsql-16beta2
+    # where the line marked with a * is the currently active version
+    #
+    # we are only interested in the first word of each line, which is the version number
+    # thus we strip the whitespace and the * from the line and split it into words
+    # and take the first word
+    versions = [line.strip('* ').split()[0] for line in output.decode('utf-8').splitlines()]
+
+    # create the list of configurations per version
+    configurations = []
+    for version in versions:
+        configurations.append(generate_configuration(version))
+
+    # create the json file
+    c_cpp_properties = {
+        "configurations": configurations,
+        "version": 4
+    }
+
+    # write the c_cpp_properties.json file
+    with open(target_path, 'w') as f:
+        json.dump(c_cpp_properties, f, indent=4)
+
+
+def generate_configuration(version):
+    """Returns a configuration for the given postgres version.
+
+    >>> generate_configuration('14.8')
+    {
+        "name": "Citus Development Configuration - Postgres 14.8",
+        "includePath": [
+            "/usr/local/include",
+            "/home/citus/.pgenv/src/postgresql-14.8/src/**",
+            "${workspaceFolder}/**",
+            "${workspaceFolder}/src/include/",
+        ],
+        "configurationProvider": "ms-vscode.makefile-tools"
+    }
+    """
+    return {
+        "name": f"Citus Development Configuration - Postgres {version}",
+        "includePath": [
+            "/usr/local/include",
+            f"/home/citus/.pgenv/src/postgresql-{version}/src/**",
+            "${workspaceFolder}/**",
+            "${workspaceFolder}/src/include/",
+        ],
+        "configurationProvider": "ms-vscode.makefile-tools"
+    }
+
+
+if __name__ == '__main__':
+    arguments = docopt(__doc__, version='0.1.0')
+    main(arguments)

.devcontainer/.vscode/launch.json

@@ -0,0 +1,40 @@
+{
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Attach Citus (devcontainer)",
+            "type": "cppdbg",
+            "request": "attach",
+            "processId": "${command:pickProcess}",
+            "program": "/home/citus/.pgenv/pgsql/bin/postgres",
+            "additionalSOLibSearchPath": "/home/citus/.pgenv/pgsql/lib",
+            "setupCommands": [
+                {
+                    "text": "handle SIGUSR1 noprint nostop pass",
+                    "description": "let gdb not stop when SIGUSR1 is sent to process",
+                    "ignoreFailures": true
+                }
+            ],
+        },
+        {
+            "name": "Open core file",
+            "type": "cppdbg",
+            "request": "launch",
+            "program": "/home/citus/.pgenv/pgsql/bin/postgres",
+            "coreDumpPath": "${input:corefile}",
+            "cwd": "${workspaceFolder}",
+            "MIMode": "gdb",
+        }
+    ],
+    "inputs": [
+        {
+            "id": "corefile",
+            "type": "command",
+            "command": "extension.commandvariable.file.pickFile",
+            "args": {
+                "dialogTitle": "Select core file",
+                "include": "**/core*",
+            },
+        },
+    ],
+}

.devcontainer/Dockerfile

@@ -0,0 +1,222 @@
+FROM ubuntu:22.04 AS base
+
+# environment is to make python pass an interactive shell, probably not the best timezone given a wide variety of colleagues
+ENV TZ=UTC
+RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
+
+# install build tools
+RUN apt update && apt install -y \
+    bison \
+    bzip2 \
+    cpanminus \
+    curl \
+    docbook-xml \
+    docbook-xsl \
+    flex \
+    gcc \
+    git \
+    libcurl4-gnutls-dev \
+    libicu-dev \
+    libkrb5-dev \
+    liblz4-dev \
+    libpam0g-dev \
+    libreadline-dev \
+    libselinux1-dev \
+    libssl-dev \
+    libxml2-utils \
+    libxslt-dev \
+    libzstd-dev \
+    locales \
+    make \
+    perl \
+    pkg-config \
+    python3 \
+    python3-pip \
+    software-properties-common \
+    sudo \
+    uuid-dev \
+    valgrind \
+    xsltproc \
+    zlib1g-dev \
+ && add-apt-repository ppa:deadsnakes/ppa -y \
+ && apt install -y \
+    python3.9-full \
+ # software properties pulls in pkexec, which makes the debugger unusable in vscode
+ && apt purge -y \
+    software-properties-common \
+ && apt autoremove -y \
+ && apt clean
+
+RUN sudo pip3 install pipenv pipenv-shebang
+
+RUN cpanm install IPC::Run
+
+RUN locale-gen en_US.UTF-8
+
+# add the citus user to sudoers and allow all sudoers to login without a password prompt
+RUN useradd -ms /bin/bash citus \
+ && usermod -aG sudo citus \
+ && echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers
+
+WORKDIR /home/citus
+USER citus
+
+# run all make commands with the number of cores available
+RUN echo "export MAKEFLAGS=\"-j \$(nproc)\"" >> "/home/citus/.bashrc"
+
+RUN git clone --branch v1.3.2 --depth 1 https://github.com/theory/pgenv.git .pgenv
+COPY --chown=citus:citus pgenv/config/ .pgenv/config/
+ENV PATH="/home/citus/.pgenv/bin:${PATH}"
+ENV PATH="/home/citus/.pgenv/pgsql/bin:${PATH}"
+
+USER citus
+
+# build postgres versions separately for effective parrallelism and caching of already built versions when changing only certain versions
+FROM base AS pg15
+RUN MAKEFLAGS="-j $(nproc)" pgenv build 15.13
+RUN rm .pgenv/src/*.tar*
+RUN make -C .pgenv/src/postgresql-*/ clean
+RUN make -C .pgenv/src/postgresql-*/src/include install
+
+# create a staging directory with all files we want to copy from our pgenv build
+# we will copy the contents of the staged folder into the final image at once
+RUN mkdir .pgenv-staging/
+RUN cp -r .pgenv/src .pgenv/pgsql-* .pgenv/config .pgenv-staging/
+RUN rm .pgenv-staging/config/default.conf
+
+FROM base AS pg16
+RUN MAKEFLAGS="-j $(nproc)" pgenv build 16.9
+RUN rm .pgenv/src/*.tar*
+RUN make -C .pgenv/src/postgresql-*/ clean
+RUN make -C .pgenv/src/postgresql-*/src/include install
+
+# create a staging directory with all files we want to copy from our pgenv build
+# we will copy the contents of the staged folder into the final image at once
+RUN mkdir .pgenv-staging/
+RUN cp -r .pgenv/src .pgenv/pgsql-* .pgenv/config .pgenv-staging/
+RUN rm .pgenv-staging/config/default.conf
+
+FROM base AS pg17
+RUN MAKEFLAGS="-j $(nproc)" pgenv build 17.5
+RUN rm .pgenv/src/*.tar*
+RUN make -C .pgenv/src/postgresql-*/ clean
+RUN make -C .pgenv/src/postgresql-*/src/include install
+
+# create a staging directory with all files we want to copy from our pgenv build
+# we will copy the contents of the staged folder into the final image at once
+RUN mkdir .pgenv-staging/
+RUN cp -r .pgenv/src .pgenv/pgsql-* .pgenv/config .pgenv-staging/
+RUN rm .pgenv-staging/config/default.conf
+
+FROM base AS uncrustify-builder
+
+RUN sudo apt update && sudo apt install -y cmake tree
+
+WORKDIR /uncrustify
+RUN curl -L https://github.com/uncrustify/uncrustify/archive/uncrustify-0.68.1.tar.gz | tar xz
+WORKDIR /uncrustify/uncrustify-uncrustify-0.68.1/
+RUN mkdir build
+WORKDIR /uncrustify/uncrustify-uncrustify-0.68.1/build/
+RUN cmake ..
+RUN MAKEFLAGS="-j $(nproc)" make -s
+
+RUN make install DESTDIR=/uncrustify
+
+# builder for all pipenv's to get them contained in a single layer
+FROM base AS pipenv
+
+WORKDIR /workspaces/citus/
+
+# tools to sync pgenv with vscode
+COPY --chown=citus:citus .vscode/Pipfile .vscode/Pipfile.lock .devcontainer/.vscode/
+RUN ( cd .devcontainer/.vscode && pipenv install )
+
+# environment to run our failure tests
+COPY --chown=citus:citus src/ src/
+RUN ( cd src/test/regress && pipenv install )
+
+# assemble the final container by copying over the artifacts from separately build containers
+FROM base AS devcontainer
+
+LABEL org.opencontainers.image.source=https://github.com/citusdata/citus
+LABEL org.opencontainers.image.description="Development container for the Citus project"
+LABEL org.opencontainers.image.licenses=AGPL-3.0-only
+
+RUN yes | sudo unminimize
+
+# install developer productivity tools
+RUN sudo apt update \
+ && sudo apt install -y \
+    autoconf2.69 \
+    bash-completion \
+    fswatch \
+    gdb \
+    htop \
+    libdbd-pg-perl \
+    libdbi-perl \
+    lsof \
+    man \
+    net-tools \
+    psmisc \
+    pspg \
+    tree \
+    vim \
+ && sudo apt clean
+
+# Since gdb will run in the context of the root user when debugging citus we will need to both
+# download the gdbpg.py script as the root user, into their home directory, as well as add .gdbinit
+# as a file owned by root
+# This will make that as soon as the debugger attaches to a postgres backend (or frankly any other process)
+# the gdbpg.py script will be sourced and the developer can direcly use it.
+RUN sudo curl -o /root/gdbpg.py https://raw.githubusercontent.com/tvesely/gdbpg/6065eee7872457785f830925eac665aa535caf62/gdbpg.py
+COPY --chown=root:root .gdbinit /root/
+
+# install developer dependencies in the global environment
+RUN --mount=type=bind,source=requirements.txt,target=requirements.txt pip install -r requirements.txt
+
+# for persistent bash history across devcontainers we need to have
+# a) a directory to store the history in
+# b) a prompt command to append the history to the file
+# c) specify the history file to store the history in
+# b and c are done in the .bashrc to make it persistent across shells only
+RUN sudo install -d -o citus -g citus /commandhistory \
+ && echo "export PROMPT_COMMAND='history -a' && export HISTFILE=/commandhistory/.bash_history" >> "/home/citus/.bashrc"
+
+# install citus-dev
+RUN git clone --branch develop https://github.com/citusdata/tools.git citus-tools \
+ && ( cd citus-tools/citus_dev && pipenv install ) \
+ && mkdir -p ~/.local/bin \
+ && ln -s /home/citus/citus-tools/citus_dev/citus_dev-pipenv .local/bin/citus_dev \
+ && sudo make -C citus-tools/uncrustify install bindir=/usr/local/bin pkgsysconfdir=/usr/local/etc/ \
+ && mkdir -p ~/.local/share/bash-completion/completions/ \
+ && ln -s ~/citus-tools/citus_dev/bash_completion ~/.local/share/bash-completion/completions/citus_dev
+
+# TODO some LC_ALL errors, possibly solved by locale-gen
+RUN git clone https://github.com/so-fancy/diff-so-fancy.git \
+ && mkdir -p ~/.local/bin \
+ && ln -s /home/citus/diff-so-fancy/diff-so-fancy .local/bin/
+
+COPY --link --from=uncrustify-builder /uncrustify/usr/ /usr/
+
+COPY --link --from=pg15 /home/citus/.pgenv-staging/ /home/citus/.pgenv/
+COPY --link --from=pg16 /home/citus/.pgenv-staging/ /home/citus/.pgenv/
+COPY --link --from=pg17 /home/citus/.pgenv-staging/ /home/citus/.pgenv/
+
+COPY --link --from=pipenv /home/citus/.local/share/virtualenvs/ /home/citus/.local/share/virtualenvs/
+
+# place to run your cluster with citus_dev
+VOLUME /data
+RUN sudo mkdir /data \
+ && sudo chown citus:citus /data
+
+COPY --chown=citus:citus .psqlrc .
+
+# with the copy linking of layers github actions seem to misbehave with the ownership of the
+# directories leading upto the link, hence a small patch layer to have to right ownerships set
+RUN sudo chown --from=root:root citus:citus -R ~
+
+# sets default pg version
+RUN pgenv switch 17.5
+
+# make connecting to the coordinator easy
+ENV PGPORT=9700

.devcontainer/Makefile

@@ -0,0 +1,11 @@
+
+init: ../.vscode/c_cpp_properties.json ../.vscode/launch.json
+
+../.vscode:
+	mkdir -p ../.vscode
+
+../.vscode/launch.json: ../.vscode .vscode/launch.json
+	cp .vscode/launch.json ../.vscode/launch.json
+
+../.vscode/c_cpp_properties.json: ../.vscode
+	./.vscode/generate_c_cpp_properties-json.py ../.vscode/c_cpp_properties.json

.devcontainer/devcontainer.json

@@ -0,0 +1,37 @@
+{
+    "image": "ghcr.io/citusdata/citus-devcontainer:main",
+    "runArgs": [
+        "--cap-add=SYS_PTRACE",
+        "--ulimit=core=-1",
+    ],
+    "forwardPorts": [
+        9700
+    ],
+    "customizations": {
+        "vscode": {
+            "extensions": [
+                "eamodio.gitlens",
+                "GitHub.copilot-chat",
+                "GitHub.copilot",
+                "github.vscode-github-actions",
+                "github.vscode-pull-request-github",
+                "ms-vscode.cpptools-extension-pack",
+                "ms-vsliveshare.vsliveshare",
+                "rioj7.command-variable",
+            ],
+            "settings": {
+                "files.exclude": {
+                    "**/*.o": true,
+                    "**/.deps/": true,
+                }
+            },
+        }
+    },
+    "mounts": [
+        "type=volume,target=/data",
+        "source=citus-bashhistory,target=/commandhistory,type=volume",
+    ],
+    "updateContentCommand": "./configure",
+    "postCreateCommand": "make -C .devcontainer/",
+}
+

.devcontainer/pgenv/config/default.conf

@@ -0,0 +1,15 @@
+PGENV_MAKE_OPTIONS=(-s)
+
+PGENV_CONFIGURE_OPTIONS=(
+    --enable-debug
+    --enable-depend
+    --enable-cassert
+    --enable-tap-tests
+    'CFLAGS=-ggdb -Og -g3 -fno-omit-frame-pointer -DUSE_VALGRIND'
+    --with-openssl
+    --with-libxml
+    --with-libxslt
+    --with-uuid=e2fs
+    --with-icu
+    --with-lz4
+)

.devcontainer/requirements.txt

@@ -0,0 +1,9 @@
+black==23.11.0
+click==8.1.7
+isort==5.12.0
+mypy-extensions==1.0.0
+packaging==23.2
+pathspec==0.11.2
+platformdirs==4.0.0
+tomli==2.0.1
+typing_extensions==4.8.0

.devcontainer/src/test/regress/Pipfile

@@ -0,0 +1,28 @@
+[[source]]
+name = "pypi"
+url = "https://pypi.python.org/simple"
+verify_ssl = true
+
+[packages]
+mitmproxy = {editable = true, ref = "main", git = "https://github.com/citusdata/mitmproxy.git"}
+construct = "*"
+docopt = "==0.6.2"
+cryptography = ">=41.0.4"
+pytest = "*"
+psycopg = "*"
+filelock = "*"
+pytest-asyncio = "*"
+pytest-timeout = "*"
+pytest-xdist = "*"
+pytest-repeat = "*"
+pyyaml = "*"
+werkzeug = "==2.3.7"
+
+[dev-packages]
+black = "*"
+isort = "*"
+flake8 = "*"
+flake8-bugbear = "*"
+
+[requires]
+python_version = "3.9"

.editorconfig

@@ -17,7 +17,7 @@ trim_trailing_whitespace = true
 insert_final_newline = unset
 trim_trailing_whitespace = unset
 
-[*.{sql,sh,py}]
+[*.{sql,sh,py,toml}]
 indent_style = space
 indent_size = 4
 tab_width = 4

.flake8

@@ -0,0 +1,7 @@
+[flake8]
+# E203 is ignored for black
+extend-ignore = E203
+# black will truncate to 88 characters usually, but long string literals it
+# might keep. That's fine in most cases unless it gets really excessive.
+max-line-length = 150
+exclude = .git,__pycache__,vendor,tmp_*

.gitattributes

@@ -25,9 +25,9 @@ configure				-whitespace
 
 # except these exceptions...
 src/backend/distributed/utils/citus_outfuncs.c -citus-style
-src/backend/distributed/deparser/ruleutils_13.c -citus-style
-src/backend/distributed/deparser/ruleutils_14.c -citus-style
 src/backend/distributed/deparser/ruleutils_15.c -citus-style
+src/backend/distributed/deparser/ruleutils_16.c -citus-style
+src/backend/distributed/deparser/ruleutils_17.c -citus-style
 src/backend/distributed/commands/index_pg_source.c -citus-style
 
 src/include/distributed/citus_nodes.h -citus-style

.github/actions/parallelization/action.yml

@@ -0,0 +1,23 @@
+name: 'Parallelization matrix'
+inputs:
+  count:
+    required: false
+    default: 32
+outputs:
+  json:
+    value: ${{ steps.generate_matrix.outputs.json }}
+runs:
+  using: "composite"
+  steps:
+    - name: Generate parallelization matrix
+      id: generate_matrix
+      shell: bash
+      run: |-
+        json_array="{\"include\": ["
+        for ((i = 1; i <= ${{ inputs.count }}; i++)); do
+            json_array+="{\"id\":\"$i\"},"
+        done
+        json_array=${json_array%,}
+        json_array+=" ]}"
+        echo "json=$json_array" >> "$GITHUB_OUTPUT"
+        echo "json=$json_array"

.github/actions/save_logs_and_results/action.yml

@@ -0,0 +1,38 @@
+name: save_logs_and_results
+inputs:
+  folder:
+    required: false
+    default: "log"
+runs:
+  using: composite
+  steps:
+  - uses: actions/upload-artifact@v4.6.0
+    name: Upload logs
+    with:
+      name: ${{ inputs.folder }}
+      if-no-files-found: ignore
+      path: |
+        src/test/**/proxy.output
+        src/test/**/results/
+        src/test/**/tmp_check/master/log
+        src/test/**/tmp_check/worker.57638/log
+        src/test/**/tmp_check/worker.57637/log
+        src/test/**/*.diffs
+        src/test/**/out/ddls.sql
+        src/test/**/out/queries.sql
+        src/test/**/logfile_*
+        /tmp/pg_upgrade_newData_logs
+  - name: Publish regression.diffs
+    run: |-
+      diffs="$(find src/test/regress -name "*.diffs" -exec cat {} \;)"
+      if ! [ -z "$diffs" ]; then
+        echo '```diff' >> $GITHUB_STEP_SUMMARY
+        echo -E "$diffs" >> $GITHUB_STEP_SUMMARY
+        echo '```' >> $GITHUB_STEP_SUMMARY
+        echo -E $diffs
+      fi
+    shell: bash
+  - name: Print stack traces
+    run: "./ci/print_stack_trace.sh"
+    if: failure()
+    shell: bash

.github/actions/setup_extension/action.yml

@@ -0,0 +1,35 @@
+name: setup_extension
+inputs:
+  pg_major:
+    required: false
+  skip_installation:
+    required: false
+    default: false
+    type: boolean
+runs:
+  using: composite
+  steps:
+  - name: Expose $PG_MAJOR to Github Env
+    run: |-
+        if [ -z "${{ inputs.pg_major }}" ]; then
+          echo "PG_MAJOR=${PG_MAJOR}" >> $GITHUB_ENV
+        else
+          echo "PG_MAJOR=${{ inputs.pg_major }}" >> $GITHUB_ENV
+        fi
+    shell: bash
+  - uses: actions/download-artifact@v4.1.8
+    with:
+      name: build-${{ env.PG_MAJOR }}
+  - name: Install Extension
+    if: ${{ inputs.skip_installation == 'false' }}
+    run: tar xfv "install-$PG_MAJOR.tar" --directory /
+    shell: bash
+  - name: Configure
+    run: |-
+      chown -R circleci .
+      git config --global --add safe.directory ${GITHUB_WORKSPACE}
+      gosu circleci ./configure --without-pg-version-check
+    shell: bash
+  - name: Enable core dumps
+    run: ulimit -c unlimited
+    shell: bash

.github/actions/upload_coverage/action.yml

@@ -0,0 +1,27 @@
+name: coverage
+inputs:
+  flags:
+    required: false
+  codecov_token:
+    required: true
+runs:
+  using: composite
+  steps:
+  - uses: codecov/codecov-action@v3
+    with:
+      flags: ${{ inputs.flags }}
+      token: ${{ inputs.codecov_token }}
+      verbose: true
+      gcov: true
+  - name: Create codeclimate coverage
+    run: |-
+      lcov --directory . --capture --output-file lcov.info
+      lcov --remove lcov.info -o lcov.info '/usr/*'
+      sed "s=^SF:$PWD/=SF:=g" -i lcov.info # relative pats are required by codeclimate
+      mkdir -p /tmp/codeclimate
+      cc-test-reporter format-coverage -t lcov -o /tmp/codeclimate/${{ inputs.flags }}.json lcov.info
+    shell: bash
+  - uses: actions/upload-artifact@v4.6.0
+    with:
+      path: "/tmp/codeclimate/*.json"
+      name: codeclimate-${{ inputs.flags }}

.github/packaging/packaging_ignore.yml

@@ -0,0 +1,3 @@
+base:
+  - ".* warning: ignoring old recipe for target [`']check'"
+  - ".* warning: overriding recipe for target [`']check'"

.github/packaging/validate_build_output.sh

@@ -0,0 +1,51 @@
+#!/bin/bash
+
+set -ex
+
+# Function to get the OS version
+get_rpm_os_version() {
+    if [[ -f /etc/centos-release ]]; then
+        cat /etc/centos-release | awk '{print $4}'
+    elif [[ -f /etc/oracle-release ]]; then
+        cat /etc/oracle-release | awk '{print $5}'
+    else
+        echo "Unknown"
+    fi
+}
+
+package_type=${1}
+
+# Since $HOME is set in GH_Actions as /github/home, pyenv fails to create virtualenvs.
+# For this script, we set $HOME to /root and then set it back to /github/home.
+GITHUB_HOME="${HOME}"
+export HOME="/root"
+
+eval "$(pyenv init -)"
+pyenv versions
+pyenv virtualenv ${PACKAGING_PYTHON_VERSION} packaging_env
+pyenv activate packaging_env
+
+git clone -b v0.8.27 --depth=1  https://github.com/citusdata/tools.git tools
+python3 -m pip install -r tools/packaging_automation/requirements.txt
+
+
+echo "Package type: ${package_type}"
+echo "OS version: $(get_rpm_os_version)"
+
+ # For RHEL 7, we need to install urllib3<2 due to below execution error
+ # ImportError: urllib3 v2.0 only supports OpenSSL 1.1.1+, currently the 'ssl'
+ # module is compiled with 'OpenSSL 1.0.2k-fips  26 Jan 2017'.
+ # See: https://github.com/urllib3/urllib3/issues/2168
+if [[ ${package_type} == "rpm" && $(get_rpm_os_version) == 7* ]]; then
+    python3 -m pip uninstall -y urllib3
+    python3 -m pip install 'urllib3<2'
+fi
+
+python3 -m tools.packaging_automation.validate_build_output --output_file output.log \
+                                                            --ignore_file .github/packaging/packaging_ignore.yml \
+                                                            --package_type ${package_type}
+pyenv deactivate
+# Set $HOME back to /github/home
+export HOME=${GITHUB_HOME}
+
+# Print the output to the console

.github/workflows/build_and_test.yml

@@ -0,0 +1,545 @@
+name: Build & Test
+run-name: Build & Test - ${{ github.event.pull_request.title || github.ref_name }}
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+on:
+  workflow_dispatch:
+    inputs:
+      skip_test_flakyness:
+        required: false
+        default: false
+        type: boolean
+  push:
+    branches:
+      - "main"
+      - "release-*"
+  pull_request:
+    types: [opened, reopened,synchronize]
+  merge_group:
+jobs:
+  # Since GHA does not interpolate env varibles in matrix context, we need to
+  # define them in a separate job and use them in other jobs.
+  params:
+    runs-on: ubuntu-latest
+    name: Initialize parameters
+    outputs:
+      build_image_name: "ghcr.io/citusdata/extbuilder"
+      test_image_name: "ghcr.io/citusdata/exttester"
+      citusupgrade_image_name: "ghcr.io/citusdata/citusupgradetester"
+      fail_test_image_name: "ghcr.io/citusdata/failtester"
+      pgupgrade_image_name: "ghcr.io/citusdata/pgupgradetester"
+      style_checker_image_name: "ghcr.io/citusdata/stylechecker"
+      style_checker_tools_version: "0.8.18"
+      sql_snapshot_pg_version: "17.5"
+      image_suffix: "-dev-d28f316"
+      pg15_version: '{ "major": "15", "full": "15.13" }'
+      pg16_version: '{ "major": "16", "full": "16.9" }'
+      pg17_version: '{ "major": "17", "full": "17.5" }'
+      upgrade_pg_versions: "15.13-16.9-17.5"
+    steps:
+      # Since GHA jobs need at least one step we use a noop step here.
+      - name: Set up parameters
+        run: echo 'noop'
+  check-sql-snapshots:
+    needs: params
+    runs-on: ubuntu-latest
+    container:
+      image: ${{ needs.params.outputs.build_image_name }}:${{ needs.params.outputs.sql_snapshot_pg_version }}${{ needs.params.outputs.image_suffix }}
+      options: --user root
+    steps:
+    - uses: actions/checkout@v4
+    - name: Check Snapshots
+      run: |
+        git config --global --add safe.directory ${GITHUB_WORKSPACE}
+        ci/check_sql_snapshots.sh
+  check-style:
+    needs: params
+    runs-on: ubuntu-latest
+    container:
+      image: ${{ needs.params.outputs.style_checker_image_name }}:${{ needs.params.outputs.style_checker_tools_version }}${{ needs.params.outputs.image_suffix }}
+    steps:
+    - name: Check Snapshots
+      run: |
+        git config --global --add safe.directory ${GITHUB_WORKSPACE}
+    - uses: actions/checkout@v4
+      with:
+        fetch-depth: 0
+    - name: Check C Style
+      run: citus_indent --check
+    - name: Check Python style
+      run: black --check .
+    - name: Check Python import order
+      run: isort --check .
+    - name: Check Python lints
+      run: flake8 .
+    - name: Fix whitespace
+      run: ci/editorconfig.sh && git diff --exit-code
+    - name: Remove useless declarations
+      run: ci/remove_useless_declarations.sh && git diff --cached --exit-code
+    - name: Sort and group includes
+      run: ci/sort_and_group_includes.sh && git diff --exit-code
+    - name: Normalize test output
+      run: ci/normalize_expected.sh && git diff --exit-code
+    - name: Check for C-style comments in migration files
+      run: ci/disallow_c_comments_in_migrations.sh && git diff --exit-code
+    - name: 'Check for comment--cached ns that start with # character in spec files'
+      run: ci/disallow_hash_comments_in_spec_files.sh && git diff --exit-code
+    - name: Check for gitignore entries .for source files
+      run: ci/fix_gitignore.sh && git diff --exit-code
+    - name: Check for lengths of changelog entries
+      run: ci/disallow_long_changelog_entries.sh
+    - name: Check for banned C API usage
+      run: ci/banned.h.sh
+    - name: Check for tests missing in schedules
+      run: ci/check_all_tests_are_run.sh
+    - name: Check if all CI scripts are actually run
+      run: ci/check_all_ci_scripts_are_run.sh
+    - name: Check if all GUCs are sorted alphabetically
+      run: ci/check_gucs_are_alphabetically_sorted.sh
+    - name: Check for missing downgrade scripts
+      run: ci/check_migration_files.sh
+  build:
+    needs: params
+    name: Build for PG${{ fromJson(matrix.pg_version).major }}
+    strategy:
+      fail-fast: false
+      matrix:
+        image_name:
+          - ${{ needs.params.outputs.build_image_name }}
+        image_suffix:
+          - ${{ needs.params.outputs.image_suffix}}
+        pg_version:
+          - ${{ needs.params.outputs.pg15_version }}
+          - ${{ needs.params.outputs.pg16_version }}
+          - ${{ needs.params.outputs.pg17_version }}
+    runs-on: ubuntu-latest
+    container:
+      image: "${{ matrix.image_name }}:${{ fromJson(matrix.pg_version).full }}${{ matrix.image_suffix }}"
+      options: --user root
+    steps:
+    - uses: actions/checkout@v4
+    - name: Expose $PG_MAJOR to Github Env
+      run: echo "PG_MAJOR=${PG_MAJOR}" >> $GITHUB_ENV
+      shell: bash
+    - name: Build
+      run: "./ci/build-citus.sh"
+      shell: bash
+    - uses: actions/upload-artifact@v4.6.0
+      with:
+        name: build-${{ env.PG_MAJOR }}
+        path: |-
+          ./build-${{ env.PG_MAJOR }}/*
+          ./install-${{ env.PG_MAJOR }}.tar
+  test-citus:
+    name: PG${{ fromJson(matrix.pg_version).major }} - ${{ matrix.make }}
+    strategy:
+      fail-fast: false
+      matrix:
+        suite:
+          - regress
+        image_name:
+          - ${{ needs.params.outputs.test_image_name }}
+        pg_version:
+          - ${{ needs.params.outputs.pg15_version }}
+          - ${{ needs.params.outputs.pg16_version }}
+          - ${{ needs.params.outputs.pg17_version }}
+        make:
+          - check-split
+          - check-multi
+          - check-multi-1
+          - check-multi-mx
+          - check-vanilla
+          - check-isolation
+          - check-operations
+          - check-follower-cluster
+          - check-columnar
+          - check-columnar-isolation
+          - check-enterprise
+          - check-enterprise-isolation
+          - check-enterprise-isolation-logicalrep-1
+          - check-enterprise-isolation-logicalrep-2
+          - check-enterprise-isolation-logicalrep-3
+        include:
+          - make: check-failure
+            pg_version: ${{ needs.params.outputs.pg15_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-failure
+            pg_version: ${{ needs.params.outputs.pg16_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-failure
+            pg_version: ${{ needs.params.outputs.pg17_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-enterprise-failure
+            pg_version: ${{ needs.params.outputs.pg15_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-enterprise-failure
+            pg_version: ${{ needs.params.outputs.pg16_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-enterprise-failure
+            pg_version: ${{ needs.params.outputs.pg17_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-pytest
+            pg_version: ${{ needs.params.outputs.pg15_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-pytest
+            pg_version: ${{ needs.params.outputs.pg16_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-pytest
+            pg_version: ${{ needs.params.outputs.pg17_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: installcheck
+            suite: cdc
+            image_name: ${{ needs.params.outputs.test_image_name }}
+            pg_version: ${{ needs.params.outputs.pg15_version }}
+          - make: installcheck
+            suite: cdc
+            image_name: ${{ needs.params.outputs.test_image_name }}
+            pg_version: ${{ needs.params.outputs.pg16_version }}
+          - make: installcheck
+            suite: cdc
+            image_name: ${{ needs.params.outputs.test_image_name }}
+            pg_version: ${{ needs.params.outputs.pg17_version }}
+          - make: check-query-generator
+            pg_version: ${{ needs.params.outputs.pg15_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-query-generator
+            pg_version: ${{ needs.params.outputs.pg16_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+          - make: check-query-generator
+            pg_version: ${{ needs.params.outputs.pg17_version }}
+            suite: regress
+            image_name: ${{ needs.params.outputs.fail_test_image_name }}
+    runs-on: ubuntu-latest
+    container:
+      image: "${{ matrix.image_name }}:${{ fromJson(matrix.pg_version).full }}${{ needs.params.outputs.image_suffix }}"
+      options: --user root --dns=8.8.8.8
+      # Due to Github creates a default network for each job, we need to use
+      # --dns= to have similar DNS settings as our other CI systems or local
+      # machines. Otherwise, we may see different results.
+    needs:
+    - params
+    - build
+    steps:
+    - uses: actions/checkout@v4
+    - uses: "./.github/actions/setup_extension"
+    - name: Run Test
+      run: gosu circleci make -C src/test/${{ matrix.suite }} ${{ matrix.make }}
+      timeout-minutes: 20
+    - uses: "./.github/actions/save_logs_and_results"
+      if: always()
+      with:
+        folder: ${{ fromJson(matrix.pg_version).major }}_${{ matrix.make }}
+    - uses: "./.github/actions/upload_coverage"
+      if: always()
+      with:
+        flags: ${{ env.PG_MAJOR }}_${{ matrix.suite }}_${{ matrix.make }}
+        codecov_token: ${{ secrets.CODECOV_TOKEN }}
+  test-arbitrary-configs:
+    name: PG${{ fromJson(matrix.pg_version).major }} - check-arbitrary-configs-${{ matrix.parallel }}
+    runs-on: ["self-hosted", "1ES.Pool=1es-gha-citusdata-pool"]
+    container:
+      image: "${{ matrix.image_name }}:${{ fromJson(matrix.pg_version).full }}${{ needs.params.outputs.image_suffix }}"
+      options: --user root
+    needs:
+      - params
+      - build
+    strategy:
+      fail-fast: false
+      matrix:
+        image_name:
+          - ${{ needs.params.outputs.fail_test_image_name }}
+        pg_version:
+          - ${{ needs.params.outputs.pg15_version }}
+          - ${{ needs.params.outputs.pg16_version }}
+          - ${{ needs.params.outputs.pg17_version }}
+        parallel: [0,1,2,3,4,5] # workaround for running 6 parallel jobs
+    steps:
+    - uses: actions/checkout@v4
+    - uses: "./.github/actions/setup_extension"
+    - name: Test arbitrary configs
+      run: |-
+        # we use parallel jobs to split the tests into 6 parts and run them in parallel
+        # the script below extracts the tests for the current job
+        N=6  # Total number of jobs (see matrix.parallel)
+        X=${{ matrix.parallel }}  # Current job number
+        TESTS=$(src/test/regress/citus_tests/print_test_names.py |
+          tr '\n' ',' | awk -v N="$N" -v X="$X" -F, '{
+            split("", parts)
+            for (i = 1; i <= NF; i++) {
+                parts[i % N] = parts[i % N] $i ","
+            }
+            print substr(parts[X], 1, length(parts[X])-1)
+        }')
+        echo $TESTS
+        gosu circleci \
+          make -C src/test/regress \
+            check-arbitrary-configs parallel=4 CONFIGS=$TESTS
+    - uses: "./.github/actions/save_logs_and_results"
+      if: always()
+      with:
+        folder: ${{ env.PG_MAJOR }}_arbitrary_configs_${{ matrix.parallel }}
+    - uses: "./.github/actions/upload_coverage"
+      if: always()
+      with:
+        flags: ${{ env.PG_MAJOR }}_arbitrary_configs_${{ matrix.parallel }}
+        codecov_token: ${{ secrets.CODECOV_TOKEN }}
+  test-pg-upgrade:
+    name: PG${{ matrix.old_pg_major }}-PG${{ matrix.new_pg_major }} - check-pg-upgrade
+    runs-on: ubuntu-latest
+    container:
+      image: "${{ needs.params.outputs.pgupgrade_image_name }}:${{ needs.params.outputs.upgrade_pg_versions }}${{ needs.params.outputs.image_suffix }}"
+      options: --user root
+    needs:
+    - params
+    - build
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - old_pg_major: 15
+            new_pg_major: 16
+          - old_pg_major: 16
+            new_pg_major: 17
+          - old_pg_major: 15
+            new_pg_major: 17
+    env:
+      old_pg_major: ${{ matrix.old_pg_major }}
+      new_pg_major: ${{ matrix.new_pg_major }}
+    steps:
+    - uses: actions/checkout@v4
+    - uses: "./.github/actions/setup_extension"
+      with:
+        pg_major: "${{ env.old_pg_major }}"
+    - uses: "./.github/actions/setup_extension"
+      with:
+        pg_major: "${{ env.new_pg_major }}"
+    - name: Install and test postgres upgrade
+      run: |-
+        gosu circleci \
+          make -C src/test/regress \
+            check-pg-upgrade \
+            old-bindir=/usr/lib/postgresql/${{ env.old_pg_major }}/bin \
+            new-bindir=/usr/lib/postgresql/${{ env.new_pg_major }}/bin
+    - name: Copy pg_upgrade logs for newData dir
+      run: |-
+        mkdir -p /tmp/pg_upgrade_newData_logs
+        if ls src/test/regress/tmp_upgrade/newData/*.log 1> /dev/null 2>&1; then
+            cp src/test/regress/tmp_upgrade/newData/*.log /tmp/pg_upgrade_newData_logs
+        fi
+      if: failure()
+    - uses: "./.github/actions/save_logs_and_results"
+      if: always()
+      with:
+        folder: ${{ env.old_pg_major }}_${{ env.new_pg_major }}_upgrade
+    - uses: "./.github/actions/upload_coverage"
+      if: always()
+      with:
+        flags: ${{ env.old_pg_major }}_${{ env.new_pg_major }}_upgrade
+        codecov_token: ${{ secrets.CODECOV_TOKEN }}
+  test-citus-upgrade:
+    name: PG${{ fromJson(needs.params.outputs.pg15_version).major }} - check-citus-upgrade
+    runs-on: ubuntu-latest
+    container:
+      image: "${{ needs.params.outputs.citusupgrade_image_name }}:${{ fromJson(needs.params.outputs.pg15_version).full }}${{ needs.params.outputs.image_suffix }}"
+      options: --user root
+    needs:
+    - params
+    - build
+    steps:
+    - uses: actions/checkout@v4
+    - uses: "./.github/actions/setup_extension"
+      with:
+        skip_installation: true
+    - name: Install and test citus upgrade
+      run: |-
+        # run make check-citus-upgrade for all citus versions
+        # the image has ${CITUS_VERSIONS} set with all verions it contains the binaries of
+        for citus_version in ${CITUS_VERSIONS}; do \
+          gosu circleci \
+            make -C src/test/regress \
+              check-citus-upgrade \
+              bindir=/usr/lib/postgresql/${PG_MAJOR}/bin \
+              citus-old-version=${citus_version} \
+              citus-pre-tar=/install-pg${PG_MAJOR}-citus${citus_version}.tar \
+              citus-post-tar=${GITHUB_WORKSPACE}/install-$PG_MAJOR.tar; \
+        done;
+        # run make check-citus-upgrade-mixed for all citus versions
+        # the image has ${CITUS_VERSIONS} set with all verions it contains the binaries of
+        for citus_version in ${CITUS_VERSIONS}; do \
+          gosu circleci \
+            make -C src/test/regress \
+              check-citus-upgrade-mixed \
+              citus-old-version=${citus_version} \
+              bindir=/usr/lib/postgresql/${PG_MAJOR}/bin \
+              citus-pre-tar=/install-pg${PG_MAJOR}-citus${citus_version}.tar \
+              citus-post-tar=${GITHUB_WORKSPACE}/install-$PG_MAJOR.tar; \
+        done;
+    - uses: "./.github/actions/save_logs_and_results"
+      if: always()
+      with:
+        folder: ${{ env.PG_MAJOR }}_citus_upgrade
+    - uses: "./.github/actions/upload_coverage"
+      if: always()
+      with:
+        flags: ${{ env.PG_MAJOR }}_citus_upgrade
+        codecov_token: ${{ secrets.CODECOV_TOKEN }}
+  upload-coverage:
+    if: always()
+    env:
+      CC_TEST_REPORTER_ID: ${{ secrets.CC_TEST_REPORTER_ID }}
+    runs-on: ubuntu-latest
+    container:
+      image: ${{ needs.params.outputs.test_image_name }}:${{ fromJson(needs.params.outputs.pg17_version).full }}${{ needs.params.outputs.image_suffix }}
+    needs:
+      - params
+      - test-citus
+      - test-arbitrary-configs
+      - test-citus-upgrade
+      - test-pg-upgrade
+    steps:
+      - uses: actions/download-artifact@v4.1.8
+        with:
+          pattern: codeclimate*
+          path: codeclimate
+          merge-multiple: true
+      - name: Upload coverage results to Code Climate
+        run: |-
+          cc-test-reporter sum-coverage codeclimate/*.json -o total.json
+          cc-test-reporter upload-coverage -i total.json
+  ch_benchmark:
+    name: CH Benchmark
+    if: startsWith(github.ref, 'refs/heads/ch_benchmark/')
+    runs-on: ubuntu-latest
+    needs:
+    - build
+    steps:
+    - uses: actions/checkout@v4
+    - uses: azure/login@v1
+      with:
+        creds: ${{ secrets.AZURE_CREDENTIALS }}
+    - name: install dependencies and run ch_benchmark tests
+      uses: azure/CLI@v1
+      with:
+        inlineScript: |
+          cd ./src/test/hammerdb
+          chmod +x run_hammerdb.sh
+          run_hammerdb.sh citusbot_ch_benchmark_rg
+  tpcc_benchmark:
+    name: TPCC Benchmark
+    if: startsWith(github.ref, 'refs/heads/tpcc_benchmark/')
+    runs-on: ubuntu-latest
+    needs:
+    - build
+    steps:
+    - uses: actions/checkout@v4
+    - uses: azure/login@v1
+      with:
+        creds: ${{ secrets.AZURE_CREDENTIALS }}
+    - name: install dependencies and run tpcc_benchmark tests
+      uses: azure/CLI@v1
+      with:
+        inlineScript: |
+          cd ./src/test/hammerdb
+          chmod +x run_hammerdb.sh
+          run_hammerdb.sh citusbot_tpcc_benchmark_rg
+  prepare_parallelization_matrix_32:
+    name: Prepare parallelization matrix
+    if: ${{ needs.test-flakyness-pre.outputs.tests != ''}}
+    needs: test-flakyness-pre
+    runs-on: ubuntu-latest
+    outputs:
+      json: ${{ steps.parallelization.outputs.json }}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: "./.github/actions/parallelization"
+        id: parallelization
+        with:
+          count: 32
+  test-flakyness-pre:
+    name: Detect regression tests need to be ran
+    if: ${{ !inputs.skip_test_flakyness }}}
+    runs-on: ubuntu-latest
+    needs: build
+    outputs:
+      tests: ${{ steps.detect-regression-tests.outputs.tests }}
+    steps:
+    - uses: actions/checkout@v4
+      with:
+        fetch-depth: 0
+    - name: Detect regression tests need to be ran
+      id: detect-regression-tests
+      run: |-
+        detected_changes=$(git diff origin/main... --name-only --diff-filter=AM | (grep 'src/test/regress/sql/.*\.sql\|src/test/regress/spec/.*\.spec\|src/test/regress/citus_tests/test/test_.*\.py' || true))
+        tests=${detected_changes}
+
+        # split the tests to be skipped --today we only skip upgrade tests
+        skipped_tests=""
+        not_skipped_tests=""
+        for test in $tests; do
+            if [[ $test =~ ^src/test/regress/sql/upgrade_ ]]; then
+                skipped_tests="$skipped_tests $test"
+            else
+                not_skipped_tests="$not_skipped_tests $test"
+            fi
+        done
+
+        if [ ! -z "$skipped_tests" ]; then
+            echo "Skipped tests " $skipped_tests
+        fi
+
+        if [ -z "$not_skipped_tests" ]; then
+            echo "Not detected any tests that flaky test detection should run"
+        else
+            echo "Detected tests " $not_skipped_tests
+        fi
+
+        echo 'tests<<EOF' >> $GITHUB_OUTPUT
+        echo "$not_skipped_tests" >> "$GITHUB_OUTPUT"
+        echo 'EOF' >> $GITHUB_OUTPUT
+  test-flakyness:
+    if: ${{ needs.test-flakyness-pre.outputs.tests != ''}}
+    name: Test flakyness
+    runs-on: ubuntu-latest
+    container:
+      image: ${{ needs.params.outputs.fail_test_image_name }}:${{ fromJson(needs.params.outputs.pg17_version).full }}${{ needs.params.outputs.image_suffix }}
+      options: --user root
+    env:
+      runs: 8
+    needs:
+    - params
+    - build
+    - test-flakyness-pre
+    - prepare_parallelization_matrix_32
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJson(needs.prepare_parallelization_matrix_32.outputs.json) }}
+    steps:
+    - uses: actions/checkout@v4
+    - uses: actions/download-artifact@v4.1.8
+    - uses: "./.github/actions/setup_extension"
+    - name: Run minimal tests
+      run: |-
+        tests="${{ needs.test-flakyness-pre.outputs.tests }}"
+        tests_array=($tests)
+        for test in "${tests_array[@]}"
+        do
+            test_name=$(echo "$test" | sed -r "s/.+\/(.+)\..+/\1/")
+            gosu circleci src/test/regress/citus_tests/run_test.py $test_name --repeat ${{ env.runs }} --use-whole-schedule-line
+        done
+      shell: bash
+    - uses: "./.github/actions/save_logs_and_results"
+      if: always()
+      with:
+        folder: test_flakyness_parallel_${{ matrix.id }}

.github/workflows/codeql.yml

@@ -0,0 +1,79 @@
+name: "CodeQL"
+
+on:
+  schedule:
+    - cron: '59 23 * * 6'
+  workflow_dispatch:
+
+jobs:
+  analyze:
+    name: Analyze
+    runs-on: ubuntu-22.04
+    permissions:
+      actions: read
+      contents: read
+      security-events: write
+
+    strategy:
+      fail-fast: false
+      matrix:
+        language: [ 'cpp', 'python']
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v4
+
+    - name: Initialize CodeQL
+      uses: github/codeql-action/init@v3
+      with:
+        languages: ${{ matrix.language }}
+
+    - name: Install package dependencies
+      run: |
+        # Create the file repository configuration:
+        sudo sh -c 'echo "deb http://apt.postgresql.org/pub/repos/apt $(lsb_release -cs)-pgdg main 15" > /etc/apt/sources.list.d/pgdg.list'
+        # Import the repository signing key:
+        wget --quiet -O - https://www.postgresql.org/media/keys/ACCC4CF8.asc | sudo apt-key add -
+        sudo apt-get update
+        sudo apt-get install -y --no-install-recommends \
+          autotools-dev \
+          build-essential \
+          ca-certificates \
+          curl \
+          debhelper \
+          devscripts \
+          fakeroot \
+          flex \
+          libcurl4-openssl-dev \
+          libdistro-info-perl \
+          libedit-dev \
+          libfile-fcntllock-perl \
+          libicu-dev \
+          libkrb5-dev \
+          liblz4-1 \
+          liblz4-dev \
+          libpam0g-dev \
+          libreadline-dev \
+          libselinux1-dev \
+          libssl-dev \
+          libxslt-dev \
+          libzstd-dev \
+          libzstd1 \
+          lintian \
+          postgresql-server-dev-15 \
+          postgresql-server-dev-all \
+          python3-pip \
+          python3-setuptools \
+          wget \
+          zlib1g-dev
+
+
+    - name: Configure, Build and Install Citus
+      if: matrix.language == 'cpp'
+      run: |
+        ./configure
+        make -sj8
+        sudo make install-all
+
+    - name: Perform CodeQL Analysis
+      uses: github/codeql-action/analyze@v3

.github/workflows/devcontainer.yml

@@ -0,0 +1,54 @@
+name: "Build devcontainer"
+
+# Since building of containers can be quite time consuming, and take up some storage,
+# there is no need to finish a build for a tag if new changes are concurrently being made.
+# This cancels any previous builds for the same tag, and only the latest one will be kept.
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+on:
+  push:
+    paths:
+      - ".devcontainer/**"
+  workflow_dispatch:
+
+jobs:
+  docker:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      packages: write
+      attestations: write
+      id-token: write
+    steps:
+      -
+        name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            ghcr.io/citusdata/citus-devcontainer
+          tags: |
+            type=ref,event=branch
+            type=sha
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v2
+      -
+        name: 'Login to GitHub Container Registry'
+        uses: docker/login-action@v3
+        with:
+          registry: ghcr.io
+          username: ${{github.actor}}
+          password: ${{secrets.GITHUB_TOKEN}}
+      -
+        name: Build and push
+        uses: docker/build-push-action@v5
+        with:
+          context: "{{defaultContext}}:.devcontainer"
+          push: true
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+          cache-from: type=gha
+          cache-to: type=gha,mode=max

.github/workflows/flaky_test_debugging.yml

@@ -0,0 +1,79 @@
+name: Flaky test debugging
+run-name: Flaky test debugging - ${{ inputs.flaky_test }} (${{ inputs.flaky_test_runs_per_job }}x${{ inputs.flaky_test_parallel_jobs }})
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+on:
+  workflow_dispatch:
+    inputs:
+      flaky_test:
+        required: true
+        type: string
+        description: Test to run
+      flaky_test_runs_per_job:
+        required: false
+        default: 8
+        type: number
+        description: Number of times to run the test
+      flaky_test_parallel_jobs:
+        required: false
+        default: 32
+        type: number
+        description: Number of parallel jobs to run
+jobs:
+  build:
+    name: Build Citus
+    runs-on: ubuntu-latest
+    container:
+      image: ${{ vars.build_image_name }}:${{ vars.pg15_version  }}${{ vars.image_suffix }}
+      options: --user root
+    steps:
+    - uses: actions/checkout@v4
+    - name: Configure, Build, and Install
+      run: |
+        echo "PG_MAJOR=${PG_MAJOR}" >> $GITHUB_ENV
+        ./ci/build-citus.sh
+      shell: bash
+    - uses: actions/upload-artifact@v4.6.0
+      with:
+        name: build-${{ env.PG_MAJOR }}
+        path: |-
+          ./build-${{ env.PG_MAJOR }}/*
+          ./install-${{ env.PG_MAJOR }}.tar
+  prepare_parallelization_matrix:
+    name: Prepare parallelization matrix
+    runs-on: ubuntu-latest
+    outputs:
+      json: ${{ steps.parallelization.outputs.json }}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: "./.github/actions/parallelization"
+        id: parallelization
+        with:
+          count: ${{ inputs.flaky_test_parallel_jobs }}
+  test_flakyness:
+    name: Test flakyness
+    runs-on: ubuntu-latest
+    container:
+      image: ${{ vars.fail_test_image_name }}:${{ vars.pg15_version  }}${{ vars.image_suffix }}
+      options: --user root
+    needs:
+      [build, prepare_parallelization_matrix]
+    env:
+      test: "${{ inputs.flaky_test }}"
+      runs: "${{ inputs.flaky_test_runs_per_job }}"
+      skip: false
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJson(needs.prepare_parallelization_matrix.outputs.json) }}
+    steps:
+    - uses: actions/checkout@v4
+    - uses: "./.github/actions/setup_extension"
+    - name: Run minimal tests
+      run: |-
+          gosu circleci src/test/regress/citus_tests/run_test.py ${{ env.test }} --repeat ${{ env.runs }} --use-whole-schedule-line
+      shell: bash
+    - uses: "./.github/actions/save_logs_and_results"
+      if: always()
+      with:
+          folder: check_flakyness_parallel_${{ matrix.id }}

.github/workflows/packaging-test-pipelines.yml

@@ -0,0 +1,177 @@
+name: Build tests in packaging images
+
+on:
+  pull_request:
+    types: [opened, reopened,synchronize]
+  merge_group:
+
+  workflow_dispatch:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+
+  get_postgres_versions_from_file:
+    runs-on: ubuntu-latest
+    outputs:
+      pg_versions: ${{ steps.get-postgres-versions.outputs.pg_versions }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 2
+      - name: Get Postgres Versions
+        id: get-postgres-versions
+        run: |
+          set -euxo pipefail
+          # Postgres versions are stored in .github/workflows/build_and_test.yml
+          # file in json strings with major and full keys.
+          # Below command extracts the versions and get the unique values.
+          pg_versions=$(cat .github/workflows/build_and_test.yml | grep -oE '"major": "[0-9]+", "full": "[0-9.]+"' | sed -E 's/"major": "([0-9]+)", "full": "([0-9.]+)"/\1/g' | sort | uniq | tr '\n', ',')
+          pg_versions_array="[ ${pg_versions} ]"
+          echo "Supported PG Versions: ${pg_versions_array}"
+          # Below line is needed to set the output variable to be used in the next job
+          echo "pg_versions=${pg_versions_array}" >> $GITHUB_OUTPUT
+        shell: bash
+  rpm_build_tests:
+    name: rpm_build_tests
+    needs: get_postgres_versions_from_file
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        # While we use separate images for different Postgres versions in rpm
+        # based distros
+        # For this reason, we need to use a "matrix" to generate names of
+        # rpm images, e.g. citus/packaging:centos-7-pg12
+        packaging_docker_image:
+          - oraclelinux-8
+          - almalinux-8
+          - almalinux-9
+        POSTGRES_VERSION: ${{ fromJson(needs.get_postgres_versions_from_file.outputs.pg_versions) }}
+
+    container:
+      image: citus/packaging:${{ matrix.packaging_docker_image }}-pg${{ matrix.POSTGRES_VERSION }}
+      options: --user root
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set Postgres and python parameters for rpm based distros
+        run: |
+          echo "/usr/pgsql-${{ matrix.POSTGRES_VERSION }}/bin" >> $GITHUB_PATH
+          echo "/root/.pyenv/bin:$PATH" >> $GITHUB_PATH
+          echo "PACKAGING_PYTHON_VERSION=3.8.16" >> $GITHUB_ENV
+
+      - name: Configure
+        run: |
+          echo "Current Shell:$0"
+          echo "GCC Version: $(gcc --version)"
+          ./configure 2>&1 | tee output.log
+
+      - name: Make clean
+        run: |
+          make clean
+
+      - name: Make
+        run: |
+          git config --global --add safe.directory ${GITHUB_WORKSPACE}
+          make CFLAGS="-Wno-missing-braces" -sj$(cat /proc/cpuinfo | grep "core id" | wc -l) 2>&1 | tee -a output.log
+
+          # Check the exit code of the make command
+          make_exit_code=${PIPESTATUS[0]}
+
+          # If the make command returned a non-zero exit code, exit with the same code
+          if [[ $make_exit_code -ne 0 ]]; then
+              echo "make command failed with exit code $make_exit_code"
+              exit $make_exit_code
+          fi
+
+      - name: Make install
+        run: |
+          make CFLAGS="-Wno-missing-braces" install 2>&1 | tee -a output.log
+
+      - name: Validate output
+        env:
+          POSTGRES_VERSION: ${{ matrix.POSTGRES_VERSION }}
+          PACKAGING_DOCKER_IMAGE: ${{ matrix.packaging_docker_image }}
+        run: |
+          echo "Postgres version: ${POSTGRES_VERSION}"
+          ./.github/packaging/validate_build_output.sh "rpm"
+
+  deb_build_tests:
+    name: deb_build_tests
+    needs: get_postgres_versions_from_file
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        # On deb based distros, we use the same docker image for
+        # builds based on different Postgres versions because deb
+        # based images include all postgres installations.
+        # For this reason, we have multiple runs --which is 3 today--
+        # for each deb based image and we use POSTGRES_VERSION to set
+        # PG_CONFIG variable in each of those runs.
+        packaging_docker_image:
+          - debian-bookworm-all
+          - debian-bullseye-all
+          - ubuntu-focal-all
+          - ubuntu-jammy-all
+
+        POSTGRES_VERSION: ${{ fromJson(needs.get_postgres_versions_from_file.outputs.pg_versions) }}
+
+    container:
+      image: citus/packaging:${{ matrix.packaging_docker_image }}
+      options: --user root
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set pg_config path and python parameters for deb based distros
+        run: |
+          echo "PG_CONFIG=/usr/lib/postgresql/${{ matrix.POSTGRES_VERSION }}/bin/pg_config" >> $GITHUB_ENV
+          echo "/root/.pyenv/bin:$PATH" >> $GITHUB_PATH
+          echo "PACKAGING_PYTHON_VERSION=3.8.16" >> $GITHUB_ENV
+
+      - name: Configure
+        run: |
+          echo "Current Shell:$0"
+          echo "GCC Version: $(gcc --version)"
+          ./configure 2>&1 | tee output.log
+
+      - name: Make clean
+        run: |
+          make clean
+
+      - name: Make
+        shell: bash
+        run: |
+          set -e
+          git config --global --add safe.directory ${GITHUB_WORKSPACE}
+          make -sj$(cat /proc/cpuinfo | grep "core id" | wc -l) 2>&1 | tee -a output.log
+
+          # Check the exit code of the make command
+          make_exit_code=${PIPESTATUS[0]}
+
+          # If the make command returned a non-zero exit code, exit with the same code
+          if [[ $make_exit_code -ne 0 ]]; then
+              echo "make command failed with exit code $make_exit_code"
+              exit $make_exit_code
+          fi
+
+
+      - name: Make install
+        run: |
+          make install 2>&1 | tee -a output.log
+
+      - name: Validate output
+        env:
+          POSTGRES_VERSION: ${{ matrix.POSTGRES_VERSION }}
+          PACKAGING_DOCKER_IMAGE: ${{ matrix.packaging_docker_image }}
+        run: |
+          echo "Postgres version: ${POSTGRES_VERSION}"
+          ./.github/packaging/validate_build_output.sh "deb"

.gitignore

@@ -38,6 +38,9 @@ lib*.pc
 /Makefile.global
 /src/Makefile.custom
 /compile_commands.json
+/src/backend/distributed/cdc/build-cdc-*/*
+/src/test/cdc/tmp_check/*
+
 
 # temporary files vim creates
 *.swp
@@ -51,3 +54,7 @@ lib*.pc
 
 # style related temporary outputs
 *.uncrustify
+.venv
+
+# added output when modifying check_gucs_are_alphabetically_sorted.sh
+guc.out

CODE_OF_CONDUCT.md

@@ -0,0 +1,9 @@
+# Microsoft Open Source Code of Conduct
+
+This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/).
+
+Resources:
+
+- [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/)
+- [Microsoft Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/)
+- Contact [opencode@microsoft.com](mailto:opencode@microsoft.com) with questions or concerns

CONTRIBUTING.md

@@ -11,6 +11,52 @@ sign a Contributor License Agreement (CLA). For an explanation of
 why we ask this as well as instructions for how to proceed, see the
 [Microsoft CLA](https://cla.opensource.microsoft.com/).
 
+### Devcontainer / Github Codespaces
+
+The easiest way to start contributing is via our devcontainer. This container works both locally in visual studio code with docker-desktop/docker-for-mac as well as [Github Codespaces](https://github.com/features/codespaces). To open the project in vscode you will need the [Dev Containers extension](https://marketplace.visualstudio.com/items?itemName=ms-vscode-remote.remote-containers). For codespaces you will need to [create a new codespace](https://codespace.new/citusdata/citus).
+
+With the extension installed you can run the following from the command pallet to get started
+
+```
+> Dev Containers: Clone Repository in Container Volume...
+```
+
+In the subsequent popup paste the url to the repo and hit enter.
+
+```
+https://github.com/citusdata/citus
+```
+
+This will create an isolated Workspace in vscode, complete with all tools required to build, test and run the Citus extension. We keep this container up to date with the supported postgres versions as well as the exact versions of tooling we use.
+
+To quickly start we suggest splitting your terminal once to have two shells. The left one in the `/workspaces/citus`, the second one changed to `/data`. The left terminal will be used to interact with the project, the right one with a testing cluster.
+
+To get citus installed from source we run `make install -s` in the first terminal. Once installed you can start a Citus cluster in the second terminal via `citus_dev make citus`. The cluster will run in the background, and can be interacted with via `citus_dev`. To get an overview of the available commands.
+
+With the Citus cluster running you can connect to the coordinator in the first terminal via `psql -p9700`. Because the coordinator is the most common entrypoint the `PGPORT` environment is set accordingly, so a simple `psql` will connect directly to the coordinator.
+
+### Debugging in the VS code
+
+1. Start Debugging: Press F5 in VS Code to start debugging. When prompted, you'll need to attach the debugger to the appropriate PostgreSQL process.
+
+2. Identify the Process: If you're running a psql command, take note of the PID that appears in your psql prompt. For example:
+```
+[local] citus@citus:9700 (PID: 5436)=#
+```
+This PID (5436 in this case) indicates the process that you should attach the debugger to.
+If you are uncertain about which process to attach, you can list all running PostgreSQL processes using the following command:
+```
+ps aux | grep postgres
+```
+
+Look for the process associated with the PID you noted. For example:
+```
+citus      5436  0.0  0.0  0  0 ?        S    14:00   0:00 postgres: citus citus
+```
+4. Attach the Debugger: Once you've identified the correct PID, select that process when prompted in VS Code to attach the debugger. You should now be able to debug the PostgreSQL session tied to the psql command.
+
+5. Set Breakpoints and Debug: With the debugger attached, you can set breakpoints within the code. This allows you to step through the code execution, inspect variables, and fully debug the PostgreSQL instance running in your container.
+
 ### Getting and building
 
 [PostgreSQL documentation](https://www.postgresql.org/support/versioning/) has a
@@ -41,6 +87,8 @@ that are missing in earlier minor versions.
 
   cd citus
   ./configure
+  # If you have already installed the project, you need to clean it first
+  make clean
   make
   make install
   # Optionally, you might instead want to use `make install-all`
@@ -79,6 +127,8 @@ that are missing in earlier minor versions.
   git clone https://github.com/citusdata/citus.git
   cd citus
   ./configure
+  # If you have already installed the project previously, you need to clean it first
+  make clean
   make
   sudo make install
   # Optionally, you might instead want to use `sudo make install-all`
@@ -129,6 +179,8 @@ that are missing in earlier minor versions.
   git clone https://github.com/citusdata/citus.git
   cd citus
   PG_CONFIG=/usr/pgsql-14/bin/pg_config ./configure
+  # If you have already installed the project previously, you need to clean it first
+  make clean
   make
   sudo make install
   # Optionally, you might instead want to use `sudo make install-all`
@@ -145,43 +197,7 @@ that are missing in earlier minor versions.
 
 ### Following our coding conventions
 
-CircleCI will automatically reject any PRs which do not follow our coding
-conventions. The easiest way to ensure your PR adheres to those conventions is
-to use the [citus_indent](https://github.com/citusdata/tools/tree/develop/uncrustify)
-tool. This tool uses `uncrustify` under the hood.
-
-```bash
-# Uncrustify changes the way it formats code every release a bit. To make sure
-# everyone formats consistently we use version 0.68.1:
-curl -L https://github.com/uncrustify/uncrustify/archive/uncrustify-0.68.1.tar.gz | tar xz
-cd uncrustify-uncrustify-0.68.1/
-mkdir build
-cd build
-cmake ..
-make -j5
-sudo make install
-cd ../..
-
-git clone https://github.com/citusdata/tools.git
-cd tools
-make uncrustify/.install
-```
-
-Once you've done that, you can run the `make reindent` command from the top
-directory to recursively check and correct the style of any source files in the
-current directory. Under the hood, `make reindent` will run `citus_indent` and
-some other style corrections for you.
-
-You can also run the following in the directory of this repository to
-automatically format all the files that you have changed before committing:
-
-```bash
-cat > .git/hooks/pre-commit << __EOF__
-#!/bin/bash
-citus_indent --check --diff || { citus_indent --diff; exit 1; }
-__EOF__
-chmod +x .git/hooks/pre-commit
-```
+Our coding conventions are documented in [STYLEGUIDE.md](STYLEGUIDE.md).
 
 ### Making SQL changes
 
@@ -218,3 +234,50 @@ style `#include` statements like this:
 
 Any other SQL you can put directly in the main sql file, e.g.
 `src/backend/distributed/sql/citus--8.3-1--9.0-1.sql`.
+
+### Backporting a commit to a release branch
+
+1. Check out the release branch that you want to backport to `git checkout release-11.3`
+2. Make sure you have the latest changes `git pull`
+3. Create a new release branch with a unique name `git checkout -b release-11.3-<yourname>`
+4. Cherry-pick the commit that you want to backport `git cherry-pick -x <sha>` (the `-x` is important)
+5. Push the branch `git push`
+6. Wait for tests to pass
+7. If the cherry-pick required non-trivial merge conflicts, create a PR and ask
+   for a review.
+8. After the tests pass on CI, fast-forward the release branch `git push origin release-11.3-<yourname>:release-11.3`
+
+### Running tests
+
+See [`src/test/regress/README.md`](https://github.com/citusdata/citus/blob/master/src/test/regress/README.md)
+
+### Documentation
+
+User-facing documentation is published on [docs.citusdata.com](https://docs.citusdata.com/). When adding a new feature, function, or setting, you can open a pull request or issue against the [Citus docs repo](https://github.com/citusdata/citus_docs/).
+
+Detailed descriptions of the implementation for Citus developers are provided in the [Citus Technical Documentation](src/backend/distributed/README.md). It is currently a single file for ease of searching. Please update the documentation if you make any changes that affect the design or add major new features.
+
+# Making a pull request ready for reviews
+
+Asking for help and asking for reviews are two different things. When you're asking for help, you're asking for someone to help you with something that you're not expected to know.
+
+But when you're asking for a review, you're asking for someone to review your work and provide feedback. So, when you're asking for a review, you're expected to make sure that:
+
+* Your changes don't perform **unnecessary line addition / deletions / style changes on unrelated files / lines**.
+
+* All CI jobs are **passing**, including **style checks** and **flaky test detection jobs**. Note that if you're an external contributor, you don't have to wait CI jobs to run (and finish) because they don't get automatically triggered for external contributors.
+
+* Your PR has necessary amount of **tests** and that they're passing.
+
+* You separated as much as possible work into **separate PRs**, e.g., a prerequisite bugfix, a refactoring etc..
+
+* Your PR doesn't introduce a typo or something that you can easily fix yourself.
+
+* After all CI jobs pass, code-coverage measurement job (CodeCov as of today) then kicks in. That's why it's important to make the **tests passing** first. At that point, you're expected to check **CodeCov annotations** that can be seen in the **Files Changed** tab and expected to make sure that it doesn't complain about any lines that are not covered. For example, it's ok if CodeCov complains about an `ereport()` call that you put for an "unexpected-but-better-than-crashing" case, but it's not ok if it complains about an uncovered `if` branch that you added.
+
+* And finally, perform a **self-review** to make sure that:
+  * Code and code-comments reflects the idea **without requiring an extra explanation** via a chat message / email / PR comment.
+    This is important because we don't expect developers to reach out to author / read about the whole discussion in the PR to understand the idea behind a commit merged into `main` branch.
+  * PR description is clear enough.
+  * If-and-only-if you're **introducing a user facing change / bugfix**, your PR has a line that starts with `DESCRIPTION: <Present simple tense word that starts with a capital letter, e.g., Adds support for / Fixes / Disallows>`.
+  * **Commit messages** are clear enough if the commits are doing logically different things.

DEVCONTAINER.md

@@ -0,0 +1,43 @@
+# Devcontainer
+
+## Coredumps
+When postgres/citus crashes, there is the option to create a coredump. This is useful for debugging the issue. Coredumps are enabled in the devcontainer by default. However, not all environments are configured correctly out of the box. The most important configuration that is not standardized is the `core_pattern`. The configuration can be verified from the container, however, you cannot change this setting from inside the container as the filesystem containing this setting is in read only mode while inside the container.
+
+To verify if corefiles are written run the following command in a terminal. This shows the filename pattern with which the corefile will be written.
+```bash
+cat /proc/sys/kernel/core_pattern
+```
+
+This should be configured with a relative path or simply a simple filename, such as `core`. When your environment shows an absolute path you will need to change this setting. How to change this setting depends highly on the underlying system as the setting needs to be changed on the kernel of the host running the container.
+
+You can put any pattern in `/proc/sys/kernel/core_pattern` as you see fit. eg. You can add the PID to the core pattern in one of two ways;
+- You either include `%p` in the core_pattern. This gets substituted with the PID of the crashing process.
+- Alternatively you could set `/proc/sys/kernel/core_uses_pid` to `1` in the same way as you set `core_pattern`. This will append the PID to the corefile if `%p` is not explicitly contained in the core_pattern.
+
+When a coredump is written you can use the debug/launch configuration `Open core file` which is preconfigured in the devcontainer. This will open a fileprompt that lists all coredumps that are found in your workspace. When you want to debug coredumps from `citus_dev` that are run in your `/data` directory, you can add the data directory to your workspace. In the command pallet of vscode you can run `>Workspace: Add Folder to Workspace...` and select the `/data` directory. This will allow you to open the coredumps from the `/data` directory in the `Open core file` debug configuration.
+
+### Windows (docker desktop)
+When running in docker desktop on windows you will most likely need to change this setting. The linux guest in WSL2 that runs your container is the `docker-desktop` environment. The easiest way to get onto the host, where you can change this setting, is to open a powershell window and verify you have the docker-desktop environment listed.
+
+```powershell
+wsl --list
+```
+
+Among others this should list both `docker-desktop` and `docker-desktop-data`. You can then open a shell in the `docker-desktop` environment.
+
+```powershell
+wsl -d docker-desktop
+```
+
+Inside this shell you can verify that you have the right environment by running
+
+```bash
+cat /proc/sys/kernel/core_pattern
+```
+
+This should show the same configuration as the one you see inside the devcontainer. You can then change the setting by running the following command.
+This will change the setting for the current session. If you want to make the change permanent you will need to add this to a startup script.
+
+```bash
+echo "core" > /proc/sys/kernel/core_pattern
+```

Makefile

@@ -11,7 +11,7 @@ endif
 
 include Makefile.global
 
-all: extension pg_send_cancellation
+all: extension
 
 
 # build columnar only
@@ -40,32 +40,28 @@ clean-full:
 
 install-downgrades:
 	$(MAKE) -C src/backend/distributed/ install-downgrades
-install-all: install-headers install-pg_send_cancellation
+install-all: install-headers
 	$(MAKE) -C src/backend/columnar/ install-all
 	$(MAKE) -C src/backend/distributed/ install-all
 
-# build citus_send_cancellation binary
-pg_send_cancellation:
-	$(MAKE) -C src/bin/pg_send_cancellation/ all
-install-pg_send_cancellation: pg_send_cancellation
-	$(MAKE) -C src/bin/pg_send_cancellation/ install
-clean-pg_send_cancellation:
-	$(MAKE) -C src/bin/pg_send_cancellation/ clean
-.PHONY: pg_send_cancellation install-pg_send_cancellation clean-pg_send_cancellation
 
 # Add to generic targets
-install: install-extension install-headers install-pg_send_cancellation
-clean: clean-extension clean-pg_send_cancellation
+install: install-extension install-headers
+clean: clean-extension
 
 # apply or check style
 reindent:
 	${citus_abs_top_srcdir}/ci/fix_style.sh
 check-style:
+	black . --check --quiet
+	isort . --check --quiet
+	flake8
 	cd ${citus_abs_top_srcdir} && citus_indent --quiet --check
 .PHONY: reindent check-style
 
 # depend on install-all so that downgrade scripts are installed as well
 check: all install-all
-	$(MAKE) -C src/test/regress check-full
+	# explicetely does not use $(MAKE) to avoid parallelism
+	make -C src/test/regress check
 
 .PHONY: all check clean install install-downgrades install-all

README.md

@@ -1,14 +1,14 @@
-| **<br/>Citus is now 100% open source and supports querying from any node.<br/><img width=1000/><br/>Read about it on the [Citus 11.0 release blog](https://www.citusdata.com/blog/2022/06/17/citus-11-goes-fully-open-source/) and the [Citus Updates page](https://www.citusdata.com/updates/).<br/><br/>** |
+| **<br/>The Citus database is 100% open source.<br/><img width=1000/><br/>Learn what's new in the [Citus 13.0 release blog](https://www.citusdata.com/blog/2025/02/06/distribute-postgresql-17-with-citus-13/) and the [Citus Updates page](https://www.citusdata.com/updates/).<br/><br/>**|
 |---|
 <br/>
 
-<br/>
 
-![Citus Banner](/citus-readme-banner.png)
+
+![Citus Banner](images/citus-readme-banner.png)
 
 [![Latest Docs](https://img.shields.io/badge/docs-latest-brightgreen.svg)](https://docs.citusdata.com/)
 [![Stack Overflow](https://img.shields.io/badge/Stack%20Overflow-%20-545353?logo=Stack%20Overflow)](https://stackoverflow.com/questions/tagged/citus)
-[![Slack Status](https://citus-slack.herokuapp.com/badge.svg)](https://citus-public.slack.com/)
+[![Slack](https://cituscdn.azureedge.net/images/social/slack-badge.svg)](https://slack.citusdata.com/)
 [![Code Coverage](https://codecov.io/gh/citusdata/citus/branch/master/graph/badge.svg)](https://app.codecov.io/gh/citusdata/citus)
 [![Twitter](https://img.shields.io/twitter/follow/citusdata.svg?label=Follow%20@citusdata)](https://twitter.com/intent/follow?screen_name=citusdata)
 
@@ -31,13 +31,15 @@ You can use these Citus superpowers to make your Postgres database scale-out rea
 
 Our [SIGMOD '21](https://2021.sigmod.org/) paper [Citus: Distributed PostgreSQL for Data-Intensive Applications](https://doi.org/10.1145/3448016.3457551) gives a more detailed look into what Citus is, how it works, and why it works that way.
 
-![Citus scales out from a single node](/citus-scale-out.png)
+![Citus scales out from a single node](images/citus-scale-out.png)
 
 Since Citus is an extension to Postgres, you can use Citus with the latest Postgres versions. And Citus works seamlessly with the PostgreSQL tools and extensions you are already familiar with.
 
 - [Why Citus?](#why-citus)
 - [Getting Started](#getting-started)
 - [Using Citus](#using-citus)
+- [Schema-based sharding](#schema-based-sharding)
+- [Setting up with High Availability](#setting-up-with-high-availability)
 - [Documentation](#documentation)
 - [Architecture](#architecture)
 - [When to Use Citus](#when-to-use-citus)
@@ -63,11 +65,11 @@ Developers choose Citus for two reasons:
 
 ## Getting Started
 
-The quickest way to get started with Citus is to use the [Hyperscale (Citus)](https://docs.microsoft.com/azure/postgresql/quickstart-create-hyperscale-portal) deployment option in the Azure Database for PostgreSQL managed service—or [set up Citus locally](https://docs.citusdata.com/en/stable/installation/single_node.html).
+The quickest way to get started with Citus is to use the [Azure Cosmos DB for PostgreSQL](https://learn.microsoft.com/azure/cosmos-db/postgresql/quickstart-create-portal) managed service in the cloud—or [set up Citus locally](https://docs.citusdata.com/en/stable/installation/single_node.html).
 
-### Hyperscale (Citus) on Azure Database for PostgreSQL
+### Citus Managed Service on Azure
 
-You can get a fully-managed Citus cluster in minutes through the Hyperscale (Citus) deployment option in the [Azure Database for PostgreSQL](https://azure.microsoft.com/services/postgresql/) portal. Azure will manage your backups, high availability through auto-failover, software updates, monitoring, and more for all of your servers. To get started with Hyperscale (Citus), use the [Hyperscale (Citus) Quickstart](https://docs.microsoft.com/azure/postgresql/quickstart-create-hyperscale-portal) in the Azure docs.
+You can get a fully-managed Citus cluster in minutes through the [Azure Cosmos DB for PostgreSQL portal](https://azure.microsoft.com/products/cosmos-db/). Azure will manage your backups, high availability through auto-failover, software updates, monitoring, and more for all of your servers. To get started Citus on Azure, use the [Azure Cosmos DB for PostgreSQL Quickstart](https://learn.microsoft.com/azure/cosmos-db/postgresql/quickstart-create-portal).
 
 ### Running Citus using Docker
 
@@ -93,14 +95,14 @@ Install packages on Ubuntu / Debian:
 ```bash
 curl https://install.citusdata.com/community/deb.sh > add-citus-repo.sh
 sudo bash add-citus-repo.sh
-sudo apt-get -y install postgresql-14-citus-11.0
+sudo apt-get -y install postgresql-17-citus-13.0
 ```
 
-Install packages on CentOS / Fedora / Red Hat:
+Install packages on Red Hat:
 ```bash
 curl https://install.citusdata.com/community/rpm.sh > add-citus-repo.sh
 sudo bash add-citus-repo.sh
-sudo yum install -y citus110_14
+sudo yum install -y citus130_17
 ```
 
 To add Citus to your local PostgreSQL database, add the following to `postgresql.conf`:
@@ -234,6 +236,41 @@ Time: 209.961 ms
 
 Co-location also helps you scale [INSERT..SELECT](https://docs.citusdata.com/en/stable/articles/aggregation.html), [stored procedures](https://www.citusdata.com/blog/2020/11/21/making-postgres-stored-procedures-9x-faster-in-citus/), and [distributed transactions](https://www.citusdata.com/blog/2017/06/02/scaling-complex-sql-transactions/).
 
+### Distributing Tables without interrupting the application
+
+
+Some of you already start with Postgres, and decide to distribute tables later on while your application using the tables. In that case, you want to avoid downtime for both reads and writes. `create_distributed_table` command block writes (e.g., DML commands) on the table until the command is finished. Instead, with `create_distributed_table_concurrently` command, your application can continue to read and write the data even during the command.
+
+
+```sql
+CREATE TABLE device_logs (
+  device_id bigint primary key,
+  log text
+);
+
+-- insert device logs
+INSERT INTO device_logs (device_id, log)
+SELECT s, 'device log:'||s FROM generate_series(0, 99) s;
+
+-- convert device_logs into a distributed table without interrupting the application
+SELECT create_distributed_table_concurrently('device_logs', 'device_id', colocate_with := 'devices');
+
+
+-- get the count of the logs, parallelized across shards
+SELECT count(*) FROM device_logs;
+
+┌───────┐
+│ count │
+├───────┤
+│   100 │
+└───────┘
+(1 row)
+
+Time: 48.734 ms
+```
+
+
+
 ### Creating Reference Tables
 
 When you need fast joins or foreign keys that do not include the distribution column, you can use `create_reference_table` to replicate a table across all nodes in the cluster.
@@ -311,9 +348,72 @@ When using columnar storage, you should only load data in batch using `COPY` or
 
 To learn more about columnar storage, check out the [columnar storage README](https://github.com/citusdata/citus/blob/master/src/backend/columnar/README.md).
 
+## Schema-based sharding
+
+Available since Citus 12.0, [schema-based sharding](https://docs.citusdata.com/en/stable/get_started/concepts.html#schema-based-sharding) is the shared database, separate schema model, the schema becomes the logical shard within the database. Multi-tenant apps can a use a schema per tenant to easily shard along the tenant dimension. Query changes are not required and the application usually only needs a small modification to set the proper search_path when switching tenants. Schema-based sharding is an ideal solution for microservices, and for ISVs deploying applications that cannot undergo the changes required to onboard row-based sharding.
+
+### Creating distributed schemas
+
+You can turn an existing schema into a distributed schema by calling `citus_schema_distribute`:
+
+```sql
+SELECT citus_schema_distribute('user_service');
+```
+
+Alternatively, you can set `citus.enable_schema_based_sharding` to have all newly created schemas be automatically converted into distributed schemas:
+
+```sql
+SET citus.enable_schema_based_sharding TO ON;
+
+CREATE SCHEMA AUTHORIZATION user_service;
+CREATE SCHEMA AUTHORIZATION time_service;
+CREATE SCHEMA AUTHORIZATION ping_service;
+```
+
+### Running queries
+
+Queries will be properly routed to schemas based on `search_path` or by explicitly using the schema name in the query.
+
+For [microservices](https://docs.citusdata.com/en/stable/get_started/tutorial_microservices.html) you would create a USER per service matching the schema name, hence the default `search_path` would contain the schema name. When connected the user queries would be automatically routed and no changes to the microservice would be required.
+
+```sql
+CREATE USER user_service;
+CREATE SCHEMA AUTHORIZATION user_service;
+```
+
+For typical multi-tenant applications, you would set the search path to the tenant schema name in your application:
+
+```sql
+SET search_path = tenant_name, public;
+```
+
+## Setting up with High Availability
+
+One of the most popular high availability solutions for PostgreSQL, [Patroni 3.0](https://github.com/zalando/patroni), has [first class support for Citus 10.0 and above](https://patroni.readthedocs.io/en/latest/citus.html#citus), additionally since Citus 11.2 ships with improvements for smoother node switchover in Patroni.
+
+An example of patronictl list output for the Citus cluster:
+
+```bash
+postgres@coord1:~$ patronictl list demo
+```
+
+```text
++ Citus cluster: demo ----------+--------------+---------+----+-----------+
+| Group | Member  | Host        | Role         | State   | TL | Lag in MB |
++-------+---------+-------------+--------------+---------+----+-----------+
+|     0 | coord1  | 172.27.0.10 | Replica      | running |  1 |         0 |
+|     0 | coord2  | 172.27.0.6  | Sync Standby | running |  1 |         0 |
+|     0 | coord3  | 172.27.0.4  | Leader       | running |  1 |           |
+|     1 | work1-1 | 172.27.0.8  | Sync Standby | running |  1 |         0 |
+|     1 | work1-2 | 172.27.0.2  | Leader       | running |  1 |           |
+|     2 | work2-1 | 172.27.0.5  | Sync Standby | running |  1 |         0 |
+|     2 | work2-2 | 172.27.0.7  | Leader       | running |  1 |           |
++-------+---------+-------------+--------------+---------+----+-----------+
+```
+
 ## Documentation
 
-If you’re ready to get started with Citus or want to know more, we recommend reading the [Citus open source documentation](https://docs.citusdata.com/en/stable/). Or, if you are using Citus on Azure, then the [Hyperscale (Citus) documentation](https://docs.microsoft.com/azure/postgresql/hyperscale/) is online and available as part of the Azure Database for PostgreSQL docs.
+If you’re ready to get started with Citus or want to know more, we recommend reading the [Citus open source documentation](https://docs.citusdata.com/en/stable/). Or, if you are using Citus on Azure, then the [Azure Cosmos DB for PostgreSQL](https://learn.microsoft.com/azure/cosmos-db/postgresql/introduction) is the place to start.
 
 Our Citus docs contain comprehensive use case guides on how to build a [multi-tenant SaaS application](https://docs.citusdata.com/en/stable/use_cases/multi_tenant.html), [real-time analytics dashboard]( https://docs.citusdata.com/en/stable/use_cases/realtime_analytics.html), or work with [time series data](https://docs.citusdata.com/en/stable/use_cases/timeseries.html).
 
@@ -323,11 +423,13 @@ A Citus database cluster grows from a single PostgreSQL node into a cluster by a
 
 Data in distributed tables is stored in “shards”, which are actually just regular PostgreSQL tables on the worker nodes. When querying a distributed table on the coordinator node, Citus will send regular SQL queries to the worker nodes. That way, all the usual PostgreSQL optimizations and extensions can automatically be used with Citus.
 
-![Citus architecture](/citus-architecture.png)
+![Citus architecture](images/citus-architecture.png)
 
 When you send a query in which all (co-located) distributed tables have the same filter on the distribution column, Citus will automatically detect that and send the whole query to the worker node that stores the data. That way, arbitrarily complex queries are supported with minimal routing overhead, which is especially useful for scaling transactional workloads. If queries do not have a specific filter, each shard is queried in parallel, which is especially useful in analytical workloads. The Citus distributed executor is adaptive and is designed to handle both query types at the same time on the same system under high concurrency, which enables large-scale mixed workloads.
 
-As of Citus 11.0, the schema and metadata of distributed tables and reference tables are automatically synchronized to all the nodes in the cluster. That way, you can connect to any node to run distributed queries. Schema changes and cluster administration still need to go through the coordinator.
+The schema and metadata of distributed tables and reference tables are automatically synchronized to all the nodes in the cluster. That way, you can connect to any node to run distributed queries. Schema changes and cluster administration still need to go through the coordinator.
+
+Detailed descriptions of the implementation for Citus developers are provided in the [Citus Technical Documentation](src/backend/distributed/README.md).
 
 ## When to use Citus
 
@@ -338,21 +440,23 @@ Citus is uniquely capable of scaling both analytical and transactional workloads
 
   The advanced parallel, distributed query engine in Citus combined with PostgreSQL features such as [array types](https://www.postgresql.org/docs/current/arrays.html), [JSONB](https://www.postgresql.org/docs/current/datatype-json.html), [lateral joins](https://heap.io/blog/engineering/postgresqls-powerful-new-join-type-lateral), and extensions like [HyperLogLog](https://github.com/citusdata/postgresql-hll) and [TopN](https://github.com/citusdata/postgresql-topn) allow you to build responsive analytics dashboards no matter how many customers or how much data you have.
 
-  Example real-time analytics users: [Algolia](https://www.citusdata.com/customers/algolia), [Heap](https://www.citusdata.com/customers/heap)
+  Example real-time analytics users: [Algolia](https://www.citusdata.com/customers/algolia)
 
 - **[Time series data](http://docs.citusdata.com/en/stable/use_cases/timeseries.html)**:
   Citus enables you to process and analyze very large amounts of time series data. The biggest Citus clusters store well over a petabyte of time series data and ingest terabytes per day.
 
   Citus integrates seamlessly with [Postgres table partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html) and has [built-in functions for partitioning by time](https://www.citusdata.com/blog/2021/10/22/how-to-scale-postgres-for-time-series-data-with-citus/), which can speed up queries and writes on time series tables. You can take advantage of Citus’s parallel, distributed query engine for fast analytical queries, and use the built-in *columnar storage* to compress old partitions.
 
-  Example users: [MixRank](https://www.citusdata.com/customers/mixrank), [Windows team](https://techcommunity.microsoft.com/t5/azure-database-for-postgresql/architecting-petabyte-scale-analytics-by-scaling-out-postgres-on/ba-p/969685)
+  Example users: [MixRank](https://www.citusdata.com/customers/mixrank)
 
 - **[Software-as-a-service (SaaS) applications](http://docs.citusdata.com/en/stable/use_cases/multi_tenant.html)**:
   SaaS and other multi-tenant applications need to be able to scale their database as the number of tenants/customers grows. Citus enables you to transparently shard a complex data model by the tenant dimension, so your database can grow along with your business.
 
   By distributing tables along a tenant ID column and co-locating data for the same tenant, Citus can horizontally scale complex (tenant-scoped) queries, transactions, and foreign key graphs. Reference tables and distributed DDL commands make database management a breeze compared to manual sharding. On top of that, you have a built-in distributed query engine for doing cross-tenant analytics inside the database.
 
-  Example multi-tenant SaaS users: [Copper](https://www.citusdata.com/customers/copper), [Salesloft](https://fivetran.com/case-studies/replicating-sharded-databases-a-case-study-of-salesloft-citus-data-and-fivetran), [ConvertFlow](https://www.citusdata.com/customers/convertflow)
+  Example multi-tenant SaaS users: [Salesloft](https://fivetran.com/case-studies/replicating-sharded-databases-a-case-study-of-salesloft-citus-data-and-fivetran), [ConvertFlow](https://www.citusdata.com/customers/convertflow)
+
+- **[Microservices](https://docs.citusdata.com/en/stable/get_started/tutorial_microservices.html)**: Citus supports schema based sharding, which allows distributing regular database schemas across many machines. This sharding methodology fits nicely with typical Microservices architecture, where storage is fully owned by the service hence can’t share the same schema definition with other tenants. Citus allows distributing horizontally scalable state across services, solving one of the [main problems](https://stackoverflow.blog/2020/11/23/the-macro-problem-with-microservices/) of microservices.
 
 - **Geospatial**:
   Because of the powerful [PostGIS](https://postgis.net/) extension to Postgres that adds support for geographic objects into Postgres, many people run spatial/GIS applications on top of Postgres. And since spatial location information has become part of our daily life, well, there are more geospatial applications than ever. When your Postgres database needs to scale out to handle an increased workload, Citus is a good fit.
@@ -365,19 +469,25 @@ Citus is uniquely capable of scaling both analytical and transactional workloads
 - **GitHub issues**: Please submit issues via [GitHub issues](https://github.com/citusdata/citus/issues).
 - **Documentation**: Our [Citus docs](https://docs.citusdata.com ) have a wealth of resources, including sections on [query performance tuning](https://docs.citusdata.com/en/stable/performance/performance_tuning.html), [useful diagnostic queries](https://docs.citusdata.com/en/stable/admin_guide/diagnostic_queries.html), and [common error messages](https://docs.citusdata.com/en/stable/reference/common_errors.html).
 - **Docs issues**: You can also submit documentation issues via [GitHub issues for our Citus docs](https://github.com/citusdata/citus_docs/issues).
-- **Updates**: Learn about what's new in each Citus version on the [Citus Updates page](https://www.citusdata.com/updates/).
+- **Updates & Release Notes**: Learn about what's new in each Citus version on the [Citus Updates page](https://www.citusdata.com/updates/).
 
 ## Contributing
 
 Citus is built on and of open source, and we welcome your contributions. The [CONTRIBUTING.md](CONTRIBUTING.md) file explains how to get started developing the Citus extension itself and our code quality guidelines.
 
+## Code of Conduct
+
+This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/).
+For more information see the [Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/) or
+contact [opencode@microsoft.com](mailto:opencode@microsoft.com) with any additional questions or comments.
+
 ## Stay Connected
 
 - **Twitter**: Follow us [@citusdata](https://twitter.com/citusdata) to track the latest posts & updates on what’s happening.
-- **Citus Blog**: Read our popular [Citus Blog](https://www.citusdata.com/blog/) for useful & informative posts about PostgreSQL and Citus.
+- **Citus Blog**: Read our popular [Citus Open Source Blog](https://www.citusdata.com/blog/) for posts about PostgreSQL and Citus.
 - **Citus Newsletter**: Subscribe to our monthly technical [Citus Newsletter](https://www.citusdata.com/join-newsletter) to get a curated collection of our favorite posts, videos, docs, talks, & other Postgres goodies.
 - **Slack**: Our [Citus Public slack](https://slack.citusdata.com/) is a good way to stay connected, not just with us but with other Citus users.
-- **Sister Blog**: Read our Azure Database for PostgreSQL [sister blog on Microsoft TechCommunity](https://techcommunity.microsoft.com/t5/azure-database-for-postgresql/bg-p/ADforPostgreSQL) for posts relating to Postgres (and Citus) on Azure.
+- **Sister Blog**: Read the PostgreSQL posts on the [Azure Cosmos DB for PostgreSQL blog](https://devblogs.microsoft.com/cosmosdb/category/postgresql/) about our managed service on Azure.
 - **Videos**: Check out this [YouTube playlist](https://www.youtube.com/playlist?list=PLixnExCn6lRq261O0iwo4ClYxHpM9qfVy) of some of our favorite Citus videos and demos. If you want to deep dive into how Citus extends PostgreSQL, you might want to check out Marco Slot’s talk at Carnegie Mellon titled [Citus: Distributed PostgreSQL as an Extension](https://youtu.be/X-aAgXJZRqM) that was part of Andy Pavlo’s Vaccination Database Talks series at CMUDB.
 - **Our other Postgres projects**: Our team also works on other awesome PostgreSQL open source extensions & projects, including: [pg_cron](https://github.com/citusdata/pg_cron), [HyperLogLog](https://github.com/citusdata/postgresql-hll), [TopN](https://github.com/citusdata/postgresql-topn), [pg_auto_failover](https://github.com/citusdata/pg_auto_failover), [activerecord-multi-tenant](https://github.com/citusdata/activerecord-multi-tenant), and [django-multitenant](https://github.com/citusdata/django-multitenant).
 

SECURITY.md

@@ -0,0 +1,41 @@
+<!-- BEGIN MICROSOFT SECURITY.MD V0.0.8 BLOCK -->
+
+## Security
+
+Microsoft takes the security of our software products and services seriously, which includes all source code repositories managed through our GitHub organizations, which include [Microsoft](https://github.com/microsoft), [Azure](https://github.com/Azure), [DotNet](https://github.com/dotnet), [AspNet](https://github.com/aspnet), [Xamarin](https://github.com/xamarin), and [our GitHub organizations](https://opensource.microsoft.com/).
+
+If you believe you have found a security vulnerability in any Microsoft-owned repository that meets [Microsoft's definition of a security vulnerability](https://aka.ms/opensource/security/definition), please report it to us as described below.
+
+## Reporting Security Issues
+
+**Please do not report security vulnerabilities through public GitHub issues.**
+
+Instead, please report them to the Microsoft Security Response Center (MSRC) at [https://msrc.microsoft.com/create-report](https://aka.ms/opensource/security/create-report).
+
+If you prefer to submit without logging in, send email to [secure@microsoft.com](mailto:secure@microsoft.com).  If possible, encrypt your message with our PGP key; please download it from the [Microsoft Security Response Center PGP Key page](https://aka.ms/opensource/security/pgpkey).
+
+You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Additional information can be found at [microsoft.com/msrc](https://aka.ms/opensource/security/msrc).
+
+Please include the requested information listed below (as much as you can provide) to help us better understand the nature and scope of the possible issue:
+
+  * Type of issue (e.g. buffer overflow, SQL injection, cross-site scripting, etc.)
+  * Full paths of source file(s) related to the manifestation of the issue
+  * The location of the affected source code (tag/branch/commit or direct URL)
+  * Any special configuration required to reproduce the issue
+  * Step-by-step instructions to reproduce the issue
+  * Proof-of-concept or exploit code (if possible)
+  * Impact of the issue, including how an attacker might exploit the issue
+
+This information will help us triage your report more quickly.
+
+If you are reporting for a bug bounty, more complete reports can contribute to a higher bounty award. Please visit our [Microsoft Bug Bounty Program](https://aka.ms/opensource/security/bounty) page for more details about our active programs.
+
+## Preferred Languages
+
+We prefer all communications to be in English.
+
+## Policy
+
+Microsoft follows the principle of [Coordinated Vulnerability Disclosure](https://aka.ms/opensource/security/cvd).
+
+<!-- END MICROSOFT SECURITY.MD BLOCK -->

STYLEGUIDE.md

@@ -0,0 +1,160 @@
+# Coding style
+
+The existing code-style in our code-base is not super consistent. There are multiple reasons for that. One big reason is because our code-base is relatively old and our standards have changed over time. The second big reason is that our style-guide is different from style-guide of Postgres and some code is copied from Postgres source code and is slightly modified. The below rules are for new code. If you're changing existing code that uses a different style, use your best judgement to decide if you use the rules here or if you match the existing style.
+
+## Using citus_indent
+
+CI pipeline will automatically reject any PRs which do not follow our coding
+conventions. The easiest way to ensure your PR adheres to those conventions is
+to use the [citus_indent](https://github.com/citusdata/tools/tree/develop/uncrustify)
+tool. This tool uses `uncrustify` under the hood.
+
+```bash
+# Uncrustify changes the way it formats code every release a bit. To make sure
+# everyone formats consistently we use version 0.68.1:
+curl -L https://github.com/uncrustify/uncrustify/archive/uncrustify-0.68.1.tar.gz | tar xz
+cd uncrustify-uncrustify-0.68.1/
+mkdir build
+cd build
+cmake ..
+make -j5
+sudo make install
+cd ../..
+
+git clone https://github.com/citusdata/tools.git
+cd tools
+make uncrustify/.install
+```
+
+Once you've done that, you can run the `make reindent` command from the top
+directory to recursively check and correct the style of any source files in the
+current directory. Under the hood, `make reindent` will run `citus_indent` and
+some other style corrections for you.
+
+You can also run the following in the directory of this repository to
+automatically format all the files that you have changed before committing:
+
+```bash
+cat > .git/hooks/pre-commit << __EOF__
+#!/bin/bash
+citus_indent --check --diff || { citus_indent --diff; exit 1; }
+__EOF__
+chmod +x .git/hooks/pre-commit
+```
+
+## Other rules we follow that citus_indent does not enforce
+
+* We almost always use **CamelCase**, when naming functions, variables etc., **not snake_case**.
+
+* We also have the habits of using a **lowerCamelCase** for some variables named from their type or from their function name, as shown in the examples:
+
+  ```c
+  bool IsCitusExtensionLoaded = false;
+
+
+  bool
+  IsAlterTableRenameStmt(RenameStmt *renameStmt)
+  {
+    AlterTableCmd *alterTableCommand = NULL;
+    ..
+    ..
+
+    bool isAlterTableRenameStmt = false;
+    ..
+  }
+  ```
+
+* We **start functions with a comment**:
+
+  ```c
+  /*
+   * MyNiceFunction <something in present simple tense, e.g., processes / returns / checks / takes X as input / does Y> ..
+   * <some more nice words> ..
+   * <some more nice words> ..
+   */
+  <static?> <return type>
+  MyNiceFunction(..)
+  {
+    ..
+    ..
+  }
+  ```
+
+* `#includes` needs to be sorted based on below ordering and then alphabetically and we should not include what we don't need in a file:
+
+  * System includes (eg. #include<...>)
+  * Postgres.h (eg. #include "postgres.h")
+  * Toplevel imports from postgres, not contained in a directory (eg. #include "miscadmin.h")
+  * General postgres includes (eg . #include "nodes/...")
+  * Toplevel citus includes, not contained in a directory (eg. #include "citus_verion.h")
+  * Columnar includes (eg. #include "columnar/...")
+  * Distributed includes (eg. #include "distributed/...")
+
+* Comments:
+  ```c
+  /* single line comments start with a lower-case */
+
+  /*
+   * We start multi-line comments with a capital letter
+   * and keep adding a star to the beginning of each line
+   * until we close the comment with a star and a slash.
+   */
+  ```
+
+* Order of function implementations and their declarations in a file:
+
+  We define static functions after the functions that call them. For example:
+
+  ```c
+  #include<..>
+  #include<..>
+  ..
+  ..
+  typedef struct
+  {
+    ..
+    ..
+  } MyNiceStruct;
+  ..
+  ..
+  PG_FUNCTION_INFO_V1(my_nice_udf1);
+  PG_FUNCTION_INFO_V1(my_nice_udf2);
+  ..
+  ..
+  // ..  somewhere on top of the file …
+  static void MyNiceStaticlyDeclaredFunction1(…);
+  static void MyNiceStaticlyDeclaredFunction2(…);
+  ..
+  ..
+
+
+  void
+  MyNiceFunctionExternedViaHeaderFile(..)
+  {
+    ..
+    ..
+    MyNiceStaticlyDeclaredFunction1(..);
+    ..
+    ..
+    MyNiceStaticlyDeclaredFunction2(..);
+    ..
+  }
+
+  ..
+  ..
+
+  // we define this first because it's called by MyNiceFunctionExternedViaHeaderFile()
+  // before MyNiceStaticlyDeclaredFunction2()
+  static void
+  MyNiceStaticlyDeclaredFunction1(…)
+  {
+  }
+  ..
+  ..
+
+  // then we define this
+  static void
+  MyNiceStaticlyDeclaredFunction2(…)
+  {
+  }
+  ```

ci/README.md

@@ -283,6 +283,14 @@ actually run in CI. This is most commonly forgotten for newly added CI tests
 that the developer only ran locally. It also checks that all CI scripts have a
 section in this `README.md` file and that they include `ci/ci_helpers.sh`.
 
+## `check_migration_files.sh`
+
+A branch that touches a set of upgrade scripts is also expected to touch
+corresponding downgrade scripts as well. If this script fails, read the output
+and make sure you update the downgrade scripts in the printed list. If you
+really don't need a downgrade to run any SQL. You can write a comment in the
+file explaining why a downgrade step is not necessary.
+
 ## `disallow_c_comments_in_migrations.sh`
 
 We do not use C-style comments in migration files as the stripped
@@ -373,3 +381,22 @@ This script checks and fixes issues with `.gitignore` rules:
 This script checks the order of the GUCs defined in `shared_library_init.c`.
 To solve this failure, please check `shared_library_init.c` and make sure that the GUC
 definitions are in alphabetical order.
+
+## `print_stack_trace.sh`
+
+This script prints stack traces for failed tests, if they left core files.
+
+## `sort_and_group_includes.sh`
+
+This script checks and fixes issues with include grouping and sorting in C files.
+
+Includes are grouped in the following groups:
+ - System includes (eg. `#include <math>`)
+ - Postgres.h include (eg. `#include "postgres.h"`)
+ - Toplevel postgres includes (includes not in a directory eg. `#include "miscadmin.h`)
+ - Postgres includes in a directory (eg. `#include "catalog/pg_type.h"`)
+ - Toplevel citus includes (includes not in a directory eg. `#include "pg_version_constants.h"`)
+ - Columnar includes (eg. `#include "columnar/columnar.h"`)
+ - Distributed includes (eg. `#include "distributed/maintenanced.h"`)
+
+Within every group the include lines are sorted alphabetically.

ci/build-citus.sh

@@ -15,9 +15,6 @@ PG_MAJOR=${PG_MAJOR:?please provide the postgres major version}
 codename=${VERSION#*(}
 codename=${codename%)*}
 
-# get project from argument
-project="${CIRCLE_PROJECT_REPONAME}"
-
 # we'll do everything with absolute paths
 basedir="$(pwd)"
 
@@ -28,7 +25,7 @@ build_ext() {
   pg_major="$1"
 
   builddir="${basedir}/build-${pg_major}"
-  echo "Beginning build of ${project} for PostgreSQL ${pg_major}..." >&2
+  echo "Beginning build for PostgreSQL ${pg_major}..." >&2
 
   # do everything in a subdirectory to avoid clutter in current directory
   mkdir -p "${builddir}" && cd "${builddir}"

ci/check_all_ci_scripts_are_run.sh

@@ -14,8 +14,8 @@ ci_scripts=$(
     grep -v -E '^(ci_helpers.sh|fix_style.sh)$'
 )
 for script in $ci_scripts; do
-    if ! grep "\\bci/$script\\b" .circleci/config.yml > /dev/null; then
-        echo "ERROR: CI script with name \"$script\" is not actually used in .circleci/config.yml"
+    if ! grep "\\bci/$script\\b" -r .github > /dev/null; then
+        echo "ERROR: CI script with name \"$script\" is not actually used in .github folder"
         exit 1
     fi
     if ! grep "^## \`$script\`\$" ci/README.md > /dev/null; then

ci/check_enterprise_merge.sh

@@ -1,96 +0,0 @@
-#!/bin/bash
-
-# Testing this script locally requires you to set the following environment
-# variables:
-# CIRCLE_BRANCH, GIT_USERNAME and GIT_TOKEN
-
-# fail if trying to reference a variable that is not set.
-set -u
-# exit immediately if a command fails
-set -e
-# Fail on pipe failures
-set -o pipefail
-
-PR_BRANCH="${CIRCLE_BRANCH}"
-ENTERPRISE_REMOTE="https://${GIT_USERNAME}:${GIT_TOKEN}@github.com/citusdata/citus-enterprise"
-
-# shellcheck disable=SC1091
-source ci/ci_helpers.sh
-
-# List executed commands. This is done so debugging this script is easier when
-# it fails. It's explicitly done after git remote add so username and password
-# are not shown in CI output (even though it's also filtered out by CircleCI)
-set -x
-
-check_compile () {
-    echo "INFO: checking if merged code can be compiled"
-    ./configure --without-libcurl
-    make -j10
-}
-
-# Clone current git repo (which should be community) to a temporary working
-# directory and go there
-GIT_DIR_ROOT="$(git rev-parse --show-toplevel)"
-TMP_GIT_DIR="$(mktemp --directory -t citus-merge-check.XXXXXXXXX)"
-git clone "$GIT_DIR_ROOT" "$TMP_GIT_DIR"
-cd "$TMP_GIT_DIR"
-
-# Fails in CI without this
-git config user.email "citus-bot@microsoft.com"
-git config user.name "citus bot"
-
-# Disable "set -x" temporarily, because $ENTERPRISE_REMOTE contains passwords
-{ set +x ; } 2> /dev/null
-git remote add enterprise "$ENTERPRISE_REMOTE"
-set -x
-
-git remote set-url --push enterprise no-pushing
-
-# Fetch enterprise-master
-git fetch enterprise enterprise-master
-
-
-git checkout "enterprise/enterprise-master"
-
-if git merge --no-commit "origin/$PR_BRANCH"; then
-    echo "INFO: community PR branch could be merged into enterprise-master"
-    # check that we can compile after the merge
-    if check_compile; then
-        exit 0
-    fi
-
-    echo "WARN: Failed to compile after community PR branch was merged into enterprise"
-fi
-
-# undo partial merge
-git merge --abort
-
-# If we have a conflict on enterprise merge on the master branch, we have a problem.
-# Provide an error message to indicate that enterprise merge is needed to fix this check.
-if [[ $PR_BRANCH = master ]]; then
-    echo "ERROR: Master branch has merge conflicts with enterprise-master."
-    echo "Try re-running this CI job after merging your changes into enterprise-master."
-    exit 1
-fi
-
-if ! git fetch enterprise "$PR_BRANCH" ; then
-    echo "ERROR: enterprise/$PR_BRANCH was not found and community PR branch could not be merged into enterprise-master"
-    exit 1
-fi
-
-# Show the top commit of the enterprise PR branch to make debugging easier
-git log -n 1 "enterprise/$PR_BRANCH"
-
-# Check that this branch contains the top commit of the current community PR
-# branch. If it does not it means it's not up to date with the current PR, so
-# the enterprise branch should be updated.
-if ! git merge-base --is-ancestor "origin/$PR_BRANCH" "enterprise/$PR_BRANCH" ; then
-    echo "ERROR: enterprise/$PR_BRANCH is not up to date with community PR branch"
-    exit 1
-fi
-
-# Now check if we can merge the enterprise PR into enterprise-master without
-# issues.
-git merge --no-commit "enterprise/$PR_BRANCH"
-# check that we can compile after the merge
-check_compile

ci/check_gucs_are_alphabetically_sorted.sh

@@ -4,7 +4,22 @@ set -euo pipefail
 # shellcheck disable=SC1091
 source ci/ci_helpers.sh
 
-# extract citus gucs in the form of "citus.X"
-grep -o -E "(\.*\"citus.\w+\")," src/backend/distributed/shared_library_init.c > gucs.out
-sort -c gucs.out
+# Find the line that exactly matches "RegisterCitusConfigVariables(void)" in
+# shared_library_init.c. grep command returns something like
+# "934:RegisterCitusConfigVariables(void)" and we extract the line number
+# with cut.
+RegisterCitusConfigVariables_begin_linenumber=$(grep -n "^RegisterCitusConfigVariables(void)$" src/backend/distributed/shared_library_init.c | cut -d: -f1)
+
+# Consider the lines starting from $RegisterCitusConfigVariables_begin_linenumber,
+# grep the first line that starts with "}" and extract the line number with cut
+# as in the previous step.
+RegisterCitusConfigVariables_length=$(tail -n +$RegisterCitusConfigVariables_begin_linenumber src/backend/distributed/shared_library_init.c | grep -n -m 1 "^}$" | cut -d: -f1)
+
+# extract the function definition of RegisterCitusConfigVariables into a temp file
+tail -n +$RegisterCitusConfigVariables_begin_linenumber src/backend/distributed/shared_library_init.c | head -n $(($RegisterCitusConfigVariables_length)) > RegisterCitusConfigVariables_func_def.out
+
+# extract citus gucs in the form of <tab><tab>"citus.X"
+grep -P "^[\t][\t]\"citus\.[a-zA-Z_0-9]+\"" RegisterCitusConfigVariables_func_def.out > gucs.out
+LC_COLLATE=C sort -c gucs.out
 rm gucs.out
+rm RegisterCitusConfigVariables_func_def.out

ci/check_migration_files.sh

@@ -0,0 +1,33 @@
+#! /bin/bash
+
+set -euo pipefail
+# shellcheck disable=SC1091
+source ci/ci_helpers.sh
+
+# This file checks for the existence of downgrade scripts for every upgrade script that is changed in the branch.
+
+# create list of migration files for upgrades
+upgrade_files=$(git diff --name-only origin/main | { grep "src/backend/distributed/sql/citus--.*sql" || exit 0 ; })
+downgrade_files=$(git diff --name-only origin/main | { grep "src/backend/distributed/sql/downgrades/citus--.*sql" || exit 0 ; })
+ret_value=0
+
+for file in $upgrade_files
+do
+    # There should always be 2 matches, and no need to avoid splitting here
+    # shellcheck disable=SC2207
+    versions=($(grep --only-matching --extended-regexp "[0-9]+\.[0-9]+[-.][0-9]+" <<< "$file"))
+
+    from_version=${versions[0]};
+    to_version=${versions[1]};
+
+    downgrade_migration_file="src/backend/distributed/sql/downgrades/citus--$to_version--$from_version.sql"
+
+    # check for the existence of migration scripts
+    if [[ $(grep --line-regexp --count "$downgrade_migration_file" <<< "$downgrade_files") == 0 ]]
+    then
+        echo "$file is updated, but $downgrade_migration_file is not updated in branch"
+        ret_value=1
+    fi
+done
+
+exit $ret_value;

ci/fix_gitignore.sh

@@ -5,7 +5,8 @@ source ci/ci_helpers.sh
 
 # Remove all the ignored files from git tree, and error out
 # find all ignored files in git tree, and use quotation marks to prevent word splitting on filenames with spaces in them
-ignored_lines_in_git_tree=$(git ls-files --ignored --exclude-standard | sed 's/.*/"&"/')
+# NOTE: Option --cached is needed to avoid a bug in git ls-files command.
+ignored_lines_in_git_tree=$(git ls-files --ignored --cached --exclude-standard | sed 's/.*/"&"/')
 
 if [[ -n $ignored_lines_in_git_tree ]]
 then

ci/fix_style.sh

@@ -9,6 +9,8 @@ cidir="${0%/*}"
 cd ${cidir}/..
 
 citus_indent . --quiet
+black . --quiet
+isort . --quiet
 ci/editorconfig.sh
 ci/remove_useless_declarations.sh
 ci/disallow_c_comments_in_migrations.sh
@@ -16,3 +18,5 @@ ci/disallow_hash_comments_in_spec_files.sh
 ci/disallow_long_changelog_entries.sh
 ci/normalize_expected.sh
 ci/fix_gitignore.sh
+ci/print_stack_trace.sh
+ci/sort_and_group_includes.sh

ci/include_grouping.py

@@ -0,0 +1,157 @@
+#!/usr/bin/env python3
+"""
+easy command line to run against all citus-style checked files:
+
+$ git ls-files \
+  | git check-attr --stdin citus-style \
+  | grep 'citus-style: set' \
+  | awk '{print $1}' \
+  | cut -d':' -f1 \
+  | xargs -n1 ./ci/include_grouping.py
+"""
+
+import collections
+import os
+import sys
+
+
+def main(args):
+    if len(args) < 2:
+        print("Usage: include_grouping.py <file>")
+        return
+
+    file = args[1]
+    if not os.path.isfile(file):
+        sys.exit(f"File '{file}' does not exist")
+
+    with open(file, "r") as in_file:
+        with open(file + ".tmp", "w") as out_file:
+            includes = []
+            skipped_lines = []
+
+            # This calls print_sorted_includes on a set of consecutive #include lines.
+            # This implicitly keeps separation of any #include lines that are contained in
+            # an #ifdef, because it will order the #include lines inside and after the
+            # #ifdef completely separately.
+            for line in in_file:
+                # if a line starts with #include we don't want to print it yet, instead we
+                # want to collect all consecutive #include lines
+                if line.startswith("#include"):
+                    includes.append(line)
+                    skipped_lines = []
+                    continue
+
+                # if we have collected any #include lines, we want to print them sorted
+                # before printing the current line. However, if the current line is empty
+                # we want to perform a lookahead to see if the next line is an #include.
+                # To maintain any separation between #include lines and their subsequent
+                # lines we keep track of all lines we have skipped inbetween.
+                if len(includes) > 0:
+                    if len(line.strip()) == 0:
+                        skipped_lines.append(line)
+                        continue
+
+                    # we have includes that need to be grouped before printing the current
+                    # line.
+                    print_sorted_includes(includes, file=out_file)
+                    includes = []
+
+                    # print any skipped lines
+                    print("".join(skipped_lines), end="", file=out_file)
+                    skipped_lines = []
+
+                print(line, end="", file=out_file)
+
+    # move out_file to file
+    os.rename(file + ".tmp", file)
+
+
+def print_sorted_includes(includes, file=sys.stdout):
+    default_group_key = 1
+    groups = collections.defaultdict(set)
+
+    # define the groups that we separate correctly. The matchers are tested in the order
+    # of their priority field. The first matcher that matches the include is used to
+    # assign the include to a group.
+    # The groups are printed in the order of their group_key.
+    matchers = [
+        {
+            "name": "system includes",
+            "matcher": lambda x: x.startswith("<"),
+            "group_key": -2,
+            "priority": 0,
+        },
+        {
+            "name": "toplevel postgres includes",
+            "matcher": lambda x: "/" not in x,
+            "group_key": 0,
+            "priority": 9,
+        },
+        {
+            "name": "postgres.h",
+            "matcher": lambda x: x.strip() in ['"postgres.h"'],
+            "group_key": -1,
+            "priority": -1,
+        },
+        {
+            "name": "toplevel citus inlcudes",
+            "matcher": lambda x: x.strip()
+            in [
+                '"citus_version.h"',
+                '"pg_version_compat.h"',
+                '"pg_version_constants.h"',
+            ],
+            "group_key": 3,
+            "priority": 0,
+        },
+        {
+            "name": "columnar includes",
+            "matcher": lambda x: x.startswith('"columnar/'),
+            "group_key": 4,
+            "priority": 1,
+        },
+        {
+            "name": "distributed includes",
+            "matcher": lambda x: x.startswith('"distributed/'),
+            "group_key": 5,
+            "priority": 1,
+        },
+    ]
+    matchers.sort(key=lambda x: x["priority"])
+
+    # throughout our codebase we have some includes where either postgres or citus
+    # includes are wrongfully included with the syntax for system includes. Before we
+    # try to match those we will change the <> to "" to make them match our system. This
+    # will also rewrite the include to the correct syntax.
+    common_system_include_error_prefixes = ["<nodes/", "<distributed/"]
+
+    # assign every include to a group
+    for include in includes:
+        # extract the group key from the include
+        include_content = include.split(" ")[1]
+
+        # fix common system includes which are secretly postgres or citus includes
+        for common_prefix in common_system_include_error_prefixes:
+            if include_content.startswith(common_prefix):
+                include_content = '"' + include_content.strip()[1:-1] + '"'
+                include = include.split(" ")[0] + " " + include_content + "\n"
+                break
+
+        group_key = default_group_key
+        for matcher in matchers:
+            if matcher["matcher"](include_content):
+                group_key = matcher["group_key"]
+                break
+
+        groups[group_key].add(include)
+
+    # iterate over all groups in the natural order of its keys
+    for i, group in enumerate(sorted(groups.items())):
+        if i > 0:
+            print(file=file)
+        includes = group[1]
+        print("".join(sorted(includes)), end="", file=file)
+
+
+if __name__ == "__main__":
+    main(sys.argv)

ci/print_stack_trace.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+set -euo pipefail
+
+# shellcheck disable=SC1091
+source ci/ci_helpers.sh
+
+# find all core files
+core_files=( $(find . -type f -regex .*core.*\d*.*postgres) )
+if [ ${#core_files[@]} -gt 0 ]; then
+    # print stack traces for the core files
+    for core_file in "${core_files[@]}"
+    do
+        # set print frame-arguments all: show all scalars + structures in the frame
+        # set print pretty on:           show structures in indented mode
+        # set print addr off:            do not show pointer address
+        # thread apply all bt full:      show stack traces for all threads
+        gdb --batch \
+            -ex "set print frame-arguments all" \
+            -ex "set print pretty on" \
+            -ex "set print addr off" \
+            -ex "thread apply all bt full" \
+            postgres "${core_file}"
+    done
+fi

ci/sort_and_group_includes.sh

@@ -0,0 +1,12 @@
+#!/bin/bash
+
+set -euo pipefail
+# shellcheck disable=SC1091
+source ci/ci_helpers.sh
+
+git ls-files \
+  | git check-attr --stdin citus-style \
+  | grep 'citus-style: set' \
+  | awk '{print $1}' \
+  | cut -d':' -f1 \
+  | xargs -n1 ./ci/include_grouping.py

configure

@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.69 for Citus 11.1devel.
+# Generated by GNU Autoconf 2.69 for Citus 13.2devel.
 #
 #
 # Copyright (C) 1992-1996, 1998-2012 Free Software Foundation, Inc.
@@ -579,8 +579,8 @@ MAKEFLAGS=
 # Identity of this package.
 PACKAGE_NAME='Citus'
 PACKAGE_TARNAME='citus'
-PACKAGE_VERSION='11.1devel'
-PACKAGE_STRING='Citus 11.1devel'
+PACKAGE_VERSION='13.2devel'
+PACKAGE_STRING='Citus 13.2devel'
 PACKAGE_BUGREPORT=''
 PACKAGE_URL=''
 
@@ -1262,7 +1262,7 @@ if test "$ac_init_help" = "long"; then
   # Omit some internal or obsolete options to make the list less imposing.
   # This message is too long to be a string in the A/UX 3.1 sh.
   cat <<_ACEOF
-\`configure' configures Citus 11.1devel to adapt to many kinds of systems.
+\`configure' configures Citus 13.2devel to adapt to many kinds of systems.
 
 Usage: $0 [OPTION]... [VAR=VALUE]...
 
@@ -1324,7 +1324,7 @@ fi
 
 if test -n "$ac_init_help"; then
   case $ac_init_help in
-     short | recursive ) echo "Configuration of Citus 11.1devel:";;
+     short | recursive ) echo "Configuration of Citus 13.2devel:";;
    esac
   cat <<\_ACEOF
 
@@ -1429,7 +1429,7 @@ fi
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
   cat <<\_ACEOF
-Citus configure 11.1devel
+Citus configure 13.2devel
 generated by GNU Autoconf 2.69
 
 Copyright (C) 2012 Free Software Foundation, Inc.
@@ -1912,7 +1912,7 @@ cat >config.log <<_ACEOF
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.
 
-It was created by Citus $as_me 11.1devel, which was
+It was created by Citus $as_me 13.2devel, which was
 generated by GNU Autoconf 2.69.  Invocation command line was
 
   $ $0 $@
@@ -2588,7 +2588,7 @@ fi
 if test "$with_pg_version_check" = no; then
     { $as_echo "$as_me:${as_lineno-$LINENO}: building against PostgreSQL $version_num (skipped compatibility check)" >&5
 $as_echo "$as_me: building against PostgreSQL $version_num (skipped compatibility check)" >&6;}
-elif test "$version_num" != '13' -a "$version_num" != '14' -a "$version_num" != '15'; then
+elif test "$version_num" != '15' -a  "$version_num" != '16' -a  "$version_num" != '17'; then
    as_fn_error $? "Citus is not compatible with the detected PostgreSQL version ${version_num}." "$LINENO" 5
 else
    { $as_echo "$as_me:${as_lineno-$LINENO}: building against PostgreSQL $version_num" >&5
@@ -5393,7 +5393,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by Citus $as_me 11.1devel, which was
+This file was extended by Citus $as_me 13.2devel, which was
 generated by GNU Autoconf 2.69.  Invocation command line was
 
   CONFIG_FILES    = $CONFIG_FILES
@@ -5455,7 +5455,7 @@ _ACEOF
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_config="`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`"
 ac_cs_version="\\
-Citus config.status 11.1devel
+Citus config.status 13.2devel
 configured by $0, generated by GNU Autoconf 2.69,
   with options \\"\$ac_cs_config\\"
 

configure.ac

@@ -5,7 +5,7 @@
 # everyone needing autoconf installed, the resulting files are checked
 # into the SCM.
 
-AC_INIT([Citus], [11.1devel])
+AC_INIT([Citus], [13.2devel])
 AC_COPYRIGHT([Copyright (c) Citus Data, Inc.])
 
 # we'll need sed and awk for some of the version commands
@@ -80,7 +80,7 @@ AC_SUBST(with_pg_version_check)
 
 if test "$with_pg_version_check" = no; then
     AC_MSG_NOTICE([building against PostgreSQL $version_num (skipped compatibility check)])
-elif test "$version_num" != '13' -a "$version_num" != '14' -a "$version_num" != '15'; then
+elif test "$version_num" != '15' -a  "$version_num" != '16' -a  "$version_num" != '17'; then
    AC_MSG_ERROR([Citus is not compatible with the detected PostgreSQL version ${version_num}.])
 else
    AC_MSG_NOTICE([building against PostgreSQL $version_num])

pyproject.toml

@@ -0,0 +1,40 @@
+[tool.isort]
+profile = 'black'
+
+[tool.black]
+include = '(src/test/regress/bin/diff-filter|\.pyi?|\.ipynb)$'
+
+[tool.pytest.ini_options]
+addopts = [
+    "--import-mode=importlib",
+    "--showlocals",
+    "--tb=short",
+]
+pythonpath = 'src/test/regress/citus_tests'
+asyncio_mode = 'auto'
+
+# Make test discovery quicker from the root dir of the repo
+testpaths = ['src/test/regress/citus_tests/test']
+
+# Make test discovery quicker from other directories than root directory
+norecursedirs = [
+    '*.egg',
+    '.*',
+    'build',
+    'venv',
+    'ci',
+    'vendor',
+    'backend',
+    'bin',
+    'include',
+    'tmp_*',
+    'results',
+    'expected',
+    'sql',
+    'spec',
+    'data',
+    '__pycache__',
+]
+
+# Don't find files with test at the end such as run_test.py
+python_files = ['test_*.py']

src/backend/columnar/.gitignore

@@ -0,0 +1,3 @@
+# The directory used to store columnar sql files after pre-processing them
+# with 'cpp' in build-time, see src/backend/columnar/Makefile.
+/build/

src/backend/columnar/Makefile

@@ -10,14 +10,51 @@ OBJS += \
 MODULE_big = citus_columnar
 EXTENSION = citus_columnar
 
-columnar_sql_files = $(patsubst $(citus_abs_srcdir)/%,%,$(wildcard $(citus_abs_srcdir)/sql/*.sql))
-columnar_downgrade_sql_files = $(patsubst $(citus_abs_srcdir)/%,%,$(wildcard $(citus_abs_srcdir)/sql/downgrades/*.sql))
-DATA = $(columnar_sql_files) \
-		$(columnar_downgrade_sql_files)
+template_sql_files = $(patsubst $(citus_abs_srcdir)/%,%,$(wildcard $(citus_abs_srcdir)/sql/*.sql))
+template_downgrade_sql_files = $(patsubst $(citus_abs_srcdir)/sql/downgrades/%,%,$(wildcard $(citus_abs_srcdir)/sql/downgrades/*.sql))
+generated_sql_files = $(patsubst %,$(citus_abs_srcdir)/build/%,$(template_sql_files))
+generated_downgrade_sql_files += $(patsubst %,$(citus_abs_srcdir)/build/sql/%,$(template_downgrade_sql_files))
+
+DATA_built = $(generated_sql_files)
 
 PG_CPPFLAGS += -I$(libpq_srcdir) -I$(safestringlib_srcdir)/include
 
 include $(citus_top_builddir)/Makefile.global
 
-.PHONY: install-all
+SQL_DEPDIR=.deps/sql
+SQL_BUILDDIR=build/sql
+
+$(generated_sql_files): $(citus_abs_srcdir)/build/%: %
+	@mkdir -p $(citus_abs_srcdir)/$(SQL_DEPDIR) $(citus_abs_srcdir)/$(SQL_BUILDDIR)
+	@# -MF is used to store dependency files(.Po) in another directory for separation
+	@# -MT is used to change the target of the rule emitted by dependency generation.
+	@# -P is used to inhibit generation of linemarkers in the output from the preprocessor.
+	@# -undef is used to not predefine any system-specific or GCC-specific macros.
+	@# `man cpp` for further information
+	cd $(citus_abs_srcdir) && cpp -undef -w -P -MMD -MP -MF$(SQL_DEPDIR)/$(*F).Po -MT$@ $< > $@
+
+$(generated_downgrade_sql_files): $(citus_abs_srcdir)/build/sql/%: sql/downgrades/%
+	@mkdir -p $(citus_abs_srcdir)/$(SQL_DEPDIR) $(citus_abs_srcdir)/$(SQL_BUILDDIR)
+	@# -MF is used to store dependency files(.Po) in another directory for separation
+	@# -MT is used to change the target of the rule emitted by dependency generation.
+	@# -P is used to inhibit generation of linemarkers in the output from the preprocessor.
+	@# -undef is used to not predefine any system-specific or GCC-specific macros.
+	@# `man cpp` for further information
+	cd $(citus_abs_srcdir) && cpp -undef -w -P -MMD -MP -MF$(SQL_DEPDIR)/$(*F).Po -MT$@ $< > $@
+
+.PHONY: install install-downgrades install-all
+
+cleanup-before-install:
+	rm -f $(DESTDIR)$(datadir)/$(datamoduledir)/citus_columnar.control
+	rm -f $(DESTDIR)$(datadir)/$(datamoduledir)/columnar--*
+	rm -f $(DESTDIR)$(datadir)/$(datamoduledir)/citus_columnar--*
+
+install: cleanup-before-install
+
+# install and install-downgrades should be run sequentially
 install-all: install
+	$(MAKE) install-downgrades
+
+install-downgrades: $(generated_downgrade_sql_files)
+	$(INSTALL_DATA) $(generated_downgrade_sql_files) '$(DESTDIR)$(datadir)/$(datamoduledir)/'
+

src/backend/columnar/README.md

@@ -52,8 +52,7 @@ Benefits of Citus Columnar over cstore_fdw:
   ... FOR UPDATE``)
 * No support for serializable isolation level
 * Support for PostgreSQL server versions 12+ only
-* No support for foreign keys, unique constraints, or exclusion
-  constraints
+* No support for foreign keys
 * No support for logical decoding
 * No support for intra-node parallel scans
 * No support for ``AFTER ... FOR EACH ROW`` triggers
@@ -234,16 +233,14 @@ CREATE TABLE perf_columnar(LIKE perf_row) USING COLUMNAR;
 ## Data
 
 ```sql
-CREATE OR REPLACE FUNCTION random_words(n INT4) RETURNS TEXT LANGUAGE plpython2u AS $$
-import random
-t = ''
-words = ['zero','one','two','three','four','five','six','seven','eight','nine','ten']
-for i in xrange(0,n):
-  if (i != 0):
-    t += ' '
-  r = random.randint(0,len(words)-1)
-  t += words[r]
-return t
+CREATE OR REPLACE FUNCTION random_words(n INT4) RETURNS TEXT LANGUAGE sql AS $$
+  WITH words(w) AS (
+    SELECT ARRAY['zero','one','two','three','four','five','six','seven','eight','nine','ten']
+  ),
+  random (word) AS (
+    SELECT w[(random()*array_length(w, 1))::int] FROM generate_series(1, $1) AS i, words
+  )
+  SELECT string_agg(word, ' ') FROM random;
 $$;
 ```
 

src/backend/columnar/citus_columnar.control

@@ -1,6 +1,6 @@
 # Columnar extension
 comment = 'Citus Columnar extension'
-default_version = '11.1-1'
+default_version = '12.2-1'
 module_pathname = '$libdir/citus_columnar'
 relocatable = false
 schema = pg_catalog

src/backend/columnar/columnar.c

@@ -11,16 +11,18 @@
  *-------------------------------------------------------------------------
  */
 
-#include "postgres.h"
-
 #include <sys/stat.h>
 #include <unistd.h>
 
+#include "postgres.h"
+
 #include "miscadmin.h"
+
 #include "utils/guc.h"
 #include "utils/rel.h"
 
 #include "citus_version.h"
+
 #include "columnar/columnar.h"
 #include "columnar/columnar_tableam.h"
 

src/backend/columnar/columnar_compression.c

@@ -13,16 +13,22 @@
  */
 #include "postgres.h"
 
-#include "citus_version.h"
 #include "common/pg_lzcompress.h"
 #include "lib/stringinfo.h"
 
+#include "citus_version.h"
+#include "pg_version_constants.h"
+
 #include "columnar/columnar_compression.h"
 
 #if HAVE_CITUS_LIBLZ4
 #include <lz4.h>
 #endif
 
+#if PG_VERSION_NUM >= PG_VERSION_16
+#include "varatt.h"
+#endif
+
 #if HAVE_LIBZSTD
 #include <zstd.h>
 #endif

src/backend/columnar/columnar_customscan.c

@@ -10,18 +10,17 @@
  *-------------------------------------------------------------------------
  */
 
-#include "citus_version.h"
+#include <math.h>
 
 #include "postgres.h"
 
-#include <math.h>
+#include "miscadmin.h"
 
 #include "access/amapi.h"
 #include "access/skey.h"
 #include "catalog/pg_am.h"
 #include "catalog/pg_statistic.h"
 #include "commands/defrem.h"
-#include "miscadmin.h"
 #include "nodes/extensible.h"
 #include "nodes/makefuncs.h"
 #include "nodes/nodeFuncs.h"
@@ -33,6 +32,10 @@
 #include "optimizer/paths.h"
 #include "optimizer/plancat.h"
 #include "optimizer/restrictinfo.h"
+#if PG_VERSION_NUM >= PG_VERSION_16
+#include "parser/parse_relation.h"
+#include "parser/parsetree.h"
+#endif
 #include "utils/builtins.h"
 #include "utils/lsyscache.h"
 #include "utils/relcache.h"
@@ -40,10 +43,13 @@
 #include "utils/selfuncs.h"
 #include "utils/spccache.h"
 
+#include "citus_version.h"
+
 #include "columnar/columnar.h"
 #include "columnar/columnar_customscan.h"
 #include "columnar/columnar_metadata.h"
 #include "columnar/columnar_tableam.h"
+
 #include "distributed/listutils.h"
 
 /*
@@ -127,6 +133,9 @@ static List * set_deparse_context_planstate(List *dpcontext, Node *node,
 /* other helpers */
 static List * ColumnarVarNeeded(ColumnarScanState *columnarScanState);
 static Bitmapset * ColumnarAttrNeeded(ScanState *ss);
+#if PG_VERSION_NUM >= PG_VERSION_16
+static Bitmapset * fixup_inherited_columns(Oid parentId, Oid childId, Bitmapset *columns);
+#endif
 
 /* saved hook value in case of unload */
 static set_rel_pathlist_hook_type PreviousSetRelPathlistHook = NULL;
@@ -198,7 +207,7 @@ columnar_customscan_init()
 		&EnableColumnarCustomScan,
 		true,
 		PGC_USERSET,
-		GUC_NO_SHOW_ALL,
+		GUC_NO_SHOW_ALL | GUC_NOT_IN_SAMPLE,
 		NULL, NULL, NULL);
 	DefineCustomBoolVariable(
 		"columnar.enable_qual_pushdown",
@@ -208,7 +217,7 @@ columnar_customscan_init()
 		&EnableColumnarQualPushdown,
 		true,
 		PGC_USERSET,
-		GUC_NO_SHOW_ALL,
+		GUC_NO_SHOW_ALL | GUC_NOT_IN_SAMPLE,
 		NULL, NULL, NULL);
 	DefineCustomRealVariable(
 		"columnar.qual_pushdown_correlation_threshold",
@@ -222,7 +231,7 @@ columnar_customscan_init()
 		0.0,
 		1.0,
 		PGC_USERSET,
-		GUC_NO_SHOW_ALL,
+		GUC_NO_SHOW_ALL | GUC_NOT_IN_SAMPLE,
 		NULL, NULL, NULL);
 	DefineCustomIntVariable(
 		"columnar.max_custom_scan_paths",
@@ -234,7 +243,7 @@ columnar_customscan_init()
 		1,
 		1024,
 		PGC_USERSET,
-		GUC_NO_SHOW_ALL,
+		GUC_NO_SHOW_ALL | GUC_NOT_IN_SAMPLE,
 		NULL, NULL, NULL);
 	DefineCustomEnumVariable(
 		"columnar.planner_debug_level",
@@ -354,7 +363,7 @@ ColumnarGetRelationInfoHook(PlannerInfo *root, Oid relationObjectId,
 
 		/* disable index-only scan */
 		IndexOptInfo *indexOptInfo = NULL;
-		foreach_ptr(indexOptInfo, rel->indexlist)
+		foreach_declared_ptr(indexOptInfo, rel->indexlist)
 		{
 			memset(indexOptInfo->canreturn, false, indexOptInfo->ncolumns * sizeof(bool));
 		}
@@ -372,7 +381,7 @@ RemovePathsByPredicate(RelOptInfo *rel, PathPredicate removePathPredicate)
 	List *filteredPathList = NIL;
 
 	Path *path = NULL;
-	foreach_ptr(path, rel->pathlist)
+	foreach_declared_ptr(path, rel->pathlist)
 	{
 		if (!removePathPredicate(path))
 		{
@@ -419,7 +428,7 @@ static void
 CostColumnarPaths(PlannerInfo *root, RelOptInfo *rel, Oid relationId)
 {
 	Path *path = NULL;
-	foreach_ptr(path, rel->pathlist)
+	foreach_declared_ptr(path, rel->pathlist)
 	{
 		if (IsA(path, IndexPath))
 		{
@@ -535,7 +544,7 @@ ColumnarIndexScanAdditionalCost(PlannerInfo *root, RelOptInfo *rel,
 	 * "anti-correlated" (-1) since both help us avoiding from reading the
 	 * same stripe again and again.
 	 */
-	double absIndexCorrelation = Abs(indexCorrelation);
+	double absIndexCorrelation = float_abs(indexCorrelation);
 
 	/*
 	 * To estimate the number of stripes that we need to read, we do linear
@@ -654,7 +663,7 @@ CheckVarStats(PlannerInfo *root, Var *var, Oid sortop, float4 *absVarCorrelation
 	 * If the Var is not highly correlated, then the chunk's min/max bounds
 	 * will be nearly useless.
 	 */
-	if (Abs(varCorrelation) < ColumnarQualPushdownCorrelationThreshold)
+	if (float_abs(varCorrelation) < ColumnarQualPushdownCorrelationThreshold)
 	{
 		if (absVarCorrelation)
 		{
@@ -662,7 +671,7 @@ CheckVarStats(PlannerInfo *root, Var *var, Oid sortop, float4 *absVarCorrelation
 			 * Report absVarCorrelation if caller wants to know why given
 			 * var is rejected.
 			 */
-			*absVarCorrelation = Abs(varCorrelation);
+			*absVarCorrelation = float_abs(varCorrelation);
 		}
 		return false;
 	}
@@ -774,7 +783,7 @@ ExtractPushdownClause(PlannerInfo *root, RelOptInfo *rel, Node *node)
 		List *pushdownableArgs = NIL;
 
 		Node *boolExprArg = NULL;
-		foreach_ptr(boolExprArg, boolExpr->args)
+		foreach_declared_ptr(boolExprArg, boolExpr->args)
 		{
 			Expr *pushdownableArg = ExtractPushdownClause(root, rel,
 														  (Node *) boolExprArg);
@@ -1042,6 +1051,15 @@ FindCandidateRelids(PlannerInfo *root, RelOptInfo *rel, List *joinClauses)
 
 	candidateRelids = bms_del_members(candidateRelids, rel->relids);
 	candidateRelids = bms_del_members(candidateRelids, rel->lateral_relids);
+
+	/*
+	 * For the relevant PG16 commit requiring this addition:
+	 * postgres/postgres@2489d76
+	 */
+#if PG_VERSION_NUM >= PG_VERSION_16
+	candidateRelids = bms_del_members(candidateRelids, root->outer_join_rels);
+#endif
+
 	return candidateRelids;
 }
 
@@ -1303,11 +1321,8 @@ AddColumnarScanPath(PlannerInfo *root, RelOptInfo *rel, RangeTblEntry *rte,
 
 	cpath->methods = &ColumnarScanPathMethods;
 
-#if (PG_VERSION_NUM >= PG_VERSION_15)
-
 	/* necessary to avoid extra Result node in PG15 */
 	cpath->flags = CUSTOMPATH_SUPPORT_PROJECTION;
-#endif
 
 	/*
 	 * populate generic path information
@@ -1371,7 +1386,43 @@ AddColumnarScanPath(PlannerInfo *root, RelOptInfo *rel, RangeTblEntry *rte,
 		cpath->custom_private = list_make2(NIL, NIL);
 	}
 
-	int numberOfColumnsRead = bms_num_members(rte->selectedCols);
+	int numberOfColumnsRead = 0;
+#if PG_VERSION_NUM >= PG_VERSION_16
+	if (rte->perminfoindex > 0)
+	{
+		/*
+		 * If perminfoindex > 0, that means that this relation's permission info
+		 * is directly found in the list of rteperminfos of the Query(root->parse)
+		 * So, all we have to do here is retrieve that info.
+		 */
+		RTEPermissionInfo *perminfo = getRTEPermissionInfo(root->parse->rteperminfos,
+														   rte);
+		numberOfColumnsRead = bms_num_members(perminfo->selectedCols);
+	}
+	else
+	{
+		/*
+		 * If perminfoindex = 0, that means we are skipping the check for permission info
+		 * for this relation, which means that it's either a partition or an inheritance child.
+		 * In these cases, we need to access the permission info of the top parent of this relation.
+		 * After thorough checking, we found that the index of the top parent pointing to the correct
+		 * range table entry in Query's range tables (root->parse->rtable) is found under
+		 * RelOptInfo rel->top_parent->relid.
+		 * For reference, check expand_partitioned_rtentry and expand_inherited_rtentry PG functions
+		 */
+		Assert(rel->top_parent);
+		RangeTblEntry *parent_rte = rt_fetch(rel->top_parent->relid, root->parse->rtable);
+		RTEPermissionInfo *perminfo = getRTEPermissionInfo(root->parse->rteperminfos,
+														   parent_rte);
+		numberOfColumnsRead = bms_num_members(fixup_inherited_columns(perminfo->relid,
+																	  rte->relid,
+																	  perminfo->
+																	  selectedCols));
+	}
+#else
+	numberOfColumnsRead = bms_num_members(rte->selectedCols);
+#endif
+
 	int numberOfClausesPushed = list_length(allClauses);
 
 	CostColumnarScan(root, rel, rte->relid, cpath, numberOfColumnsRead,
@@ -1391,6 +1442,69 @@ AddColumnarScanPath(PlannerInfo *root, RelOptInfo *rel, RangeTblEntry *rte,
 }
 
 
+#if PG_VERSION_NUM >= PG_VERSION_16
+
+/*
+ * fixup_inherited_columns
+ *
+ * Exact function Copied from PG16 as it's static.
+ *
+ * When user is querying on a table with children, it implicitly accesses
+ * child tables also. So, we also need to check security label of child
+ * tables and columns, but there is no guarantee attribute numbers are
+ * same between the parent and children.
+ * It returns a bitmapset which contains attribute number of the child
+ * table based on the given bitmapset of the parent.
+ */
+static Bitmapset *
+fixup_inherited_columns(Oid parentId, Oid childId, Bitmapset *columns)
+{
+	Bitmapset *result = NULL;
+
+	/*
+	 * obviously, no need to do anything here
+	 */
+	if (parentId == childId)
+	{
+		return columns;
+	}
+
+	int index = -1;
+	while ((index = bms_next_member(columns, index)) >= 0)
+	{
+		/* bit numbers are offset by FirstLowInvalidHeapAttributeNumber */
+		AttrNumber attno = index + FirstLowInvalidHeapAttributeNumber;
+
+		/*
+		 * whole-row-reference shall be fixed-up later
+		 */
+		if (attno == InvalidAttrNumber)
+		{
+			result = bms_add_member(result, index);
+			continue;
+		}
+
+		char *attname = get_attname(parentId, attno, false);
+		attno = get_attnum(childId, attname);
+		if (attno == InvalidAttrNumber)
+		{
+			elog(ERROR, "cache lookup failed for attribute %s of relation %u",
+				 attname, childId);
+		}
+
+		result = bms_add_member(result,
+								attno - FirstLowInvalidHeapAttributeNumber);
+
+		pfree(attname);
+	}
+
+	return result;
+}
+
+
+#endif
+
+
 /*
  * CostColumnarScan calculates the cost of scanning the columnar table. The
  * cost is estimated by using all stripe metadata to estimate based on the
@@ -1435,13 +1549,14 @@ ColumnarPerStripeScanCost(RelOptInfo *rel, Oid relationId, int numberOfColumnsRe
 		ereport(ERROR, (errmsg("could not open relation with OID %u", relationId)));
 	}
 
-	List *stripeList = StripesForRelfilenode(relation->rd_node);
+	List *stripeList = StripesForRelfilelocator(RelationPhysicalIdentifier_compat(
+													relation));
 	RelationClose(relation);
 
 	uint32 maxColumnCount = 0;
 	uint64 totalStripeSize = 0;
 	StripeMetadata *stripeMetadata = NULL;
-	foreach_ptr(stripeMetadata, stripeList)
+	foreach_declared_ptr(stripeMetadata, stripeList)
 	{
 		totalStripeSize += stripeMetadata->dataLength;
 		maxColumnCount = Max(maxColumnCount, stripeMetadata->columnCount);
@@ -1492,7 +1607,8 @@ ColumnarTableStripeCount(Oid relationId)
 		ereport(ERROR, (errmsg("could not open relation with OID %u", relationId)));
 	}
 
-	List *stripeList = StripesForRelfilenode(relation->rd_node);
+	List *stripeList = StripesForRelfilelocator(RelationPhysicalIdentifier_compat(
+													relation));
 	int stripeCount = list_length(stripeList);
 	RelationClose(relation);
 
@@ -1814,11 +1930,6 @@ ColumnarScan_EndCustomScan(CustomScanState *node)
 	 */
 	TableScanDesc scanDesc = node->ss.ss_currentScanDesc;
 
-	/*
-	 * Free the exprcontext
-	 */
-	ExecFreeExprContext(&node->ss.ps);
-
 	/*
 	 * clean out the tuple table
 	 */

src/backend/columnar/columnar_debug.c

@@ -11,12 +11,12 @@
 #include "postgres.h"
 
 #include "funcapi.h"
+#include "miscadmin.h"
+
 #include "access/nbtree.h"
 #include "access/table.h"
 #include "catalog/pg_am.h"
 #include "catalog/pg_type.h"
-#include "distributed/pg_version_constants.h"
-#include "miscadmin.h"
 #include "storage/fd.h"
 #include "storage/smgr.h"
 #include "utils/guc.h"
@@ -25,6 +25,8 @@
 #include "utils/tuplestore.h"
 
 #include "pg_version_compat.h"
+#include "pg_version_constants.h"
+
 #include "columnar/columnar.h"
 #include "columnar/columnar_storage.h"
 #include "columnar/columnar_version_compat.h"
@@ -159,5 +161,5 @@ MemoryContextTotals(MemoryContext context, MemoryContextCounters *counters)
 		MemoryContextTotals(child, counters);
 	}
 
-	context->methods->stats_compat(context, NULL, NULL, counters, true);
+	context->methods->stats(context, NULL, NULL, counters, true);
 }

src/backend/columnar/columnar_metadata.c

@@ -19,47 +19,62 @@
  */
 
 
+#include <sys/stat.h>
+
 #include "postgres.h"
 
+#include "miscadmin.h"
+#include "port.h"
 #include "safe_lib.h"
 
-#include "citus_version.h"
-#include "columnar/columnar.h"
-#include "columnar/columnar_storage.h"
-#include "columnar/columnar_version_compat.h"
-#include "distributed/listutils.h"
-
-#include <sys/stat.h>
 #include "access/heapam.h"
 #include "access/htup_details.h"
 #include "access/nbtree.h"
 #include "access/xact.h"
 #include "catalog/indexing.h"
-#include "catalog/pg_namespace.h"
-#include "catalog/pg_collation.h"
-#include "catalog/pg_type.h"
 #include "catalog/namespace.h"
+#include "catalog/pg_collation.h"
+#include "catalog/pg_namespace.h"
+#include "catalog/pg_type.h"
 #include "commands/defrem.h"
 #include "commands/sequence.h"
 #include "commands/trigger.h"
 #include "executor/executor.h"
 #include "executor/spi.h"
-#include "miscadmin.h"
-#include "nodes/execnodes.h"
 #include "lib/stringinfo.h"
-#include "port.h"
+#include "nodes/execnodes.h"
 #include "storage/fd.h"
 #include "storage/lmgr.h"
 #include "storage/procarray.h"
 #include "storage/smgr.h"
 #include "utils/builtins.h"
 #include "utils/fmgroids.h"
-#include "utils/memutils.h"
 #include "utils/lsyscache.h"
+#include "utils/memutils.h"
 #include "utils/rel.h"
+
+#include "citus_version.h"
+#include "pg_version_constants.h"
+
+#include "columnar/columnar.h"
+#include "columnar/columnar_storage.h"
+#include "columnar/columnar_version_compat.h"
+
+#include "distributed/listutils.h"
+
+#if PG_VERSION_NUM >= PG_VERSION_16
+#include "parser/parse_relation.h"
+#include "storage/relfilelocator.h"
+#include "utils/relfilenumbermap.h"
+#else
 #include "utils/relfilenodemap.h"
+#endif
 
 #define COLUMNAR_RELOPTION_NAMESPACE "columnar"
+#define SLOW_METADATA_ACCESS_WARNING \
+	"Metadata index %s is not available, this might mean slower read/writes " \
+	"on columnar tables. This is expected during Postgres upgrades and not " \
+	"expected otherwise."
 
 typedef struct
 {
@@ -108,7 +123,7 @@ static Oid ColumnarChunkGroupRelationId(void);
 static Oid ColumnarChunkIndexRelationId(void);
 static Oid ColumnarChunkGroupIndexRelationId(void);
 static Oid ColumnarNamespaceId(void);
-static uint64 LookupStorageId(RelFileNode relfilenode);
+static uint64 LookupStorageId(RelFileLocator relfilelocator);
 static uint64 GetHighestUsedRowNumber(uint64 storageId);
 static void DeleteStorageFromColumnarMetadataTable(Oid metadataTableId,
 												   AttrNumber storageIdAtrrNumber,
@@ -587,14 +602,15 @@ ReadColumnarOptions(Oid regclass, ColumnarOptions *options)
  * of columnar.chunk.
  */
 void
-SaveStripeSkipList(RelFileNode relfilenode, uint64 stripe, StripeSkipList *chunkList,
+SaveStripeSkipList(RelFileLocator relfilelocator, uint64 stripe,
+				   StripeSkipList *chunkList,
 				   TupleDesc tupleDescriptor)
 {
 	uint32 columnIndex = 0;
 	uint32 chunkIndex = 0;
 	uint32 columnCount = chunkList->columnCount;
 
-	uint64 storageId = LookupStorageId(relfilenode);
+	uint64 storageId = LookupStorageId(relfilelocator);
 	Oid columnarChunkOid = ColumnarChunkRelationId();
 	Relation columnarChunk = table_open(columnarChunkOid, RowExclusiveLock);
 	ModifyState *modifyState = StartModifyRelation(columnarChunk);
@@ -653,10 +669,10 @@ SaveStripeSkipList(RelFileNode relfilenode, uint64 stripe, StripeSkipList *chunk
  * SaveChunkGroups saves the metadata for given chunk groups in columnar.chunk_group.
  */
 void
-SaveChunkGroups(RelFileNode relfilenode, uint64 stripe,
+SaveChunkGroups(RelFileLocator relfilelocator, uint64 stripe,
 				List *chunkGroupRowCounts)
 {
-	uint64 storageId = LookupStorageId(relfilenode);
+	uint64 storageId = LookupStorageId(relfilelocator);
 	Oid columnarChunkGroupOid = ColumnarChunkGroupRelationId();
 	Relation columnarChunkGroup = table_open(columnarChunkGroupOid, RowExclusiveLock);
 	ModifyState *modifyState = StartModifyRelation(columnarChunkGroup);
@@ -689,7 +705,8 @@ SaveChunkGroups(RelFileNode relfilenode, uint64 stripe,
  * ReadStripeSkipList fetches chunk metadata for a given stripe.
  */
 StripeSkipList *
-ReadStripeSkipList(RelFileNode relfilenode, uint64 stripe, TupleDesc tupleDescriptor,
+ReadStripeSkipList(RelFileLocator relfilelocator, uint64 stripe,
+				   TupleDesc tupleDescriptor,
 				   uint32 chunkCount, Snapshot snapshot)
 {
 	int32 columnIndex = 0;
@@ -697,19 +714,27 @@ ReadStripeSkipList(RelFileNode relfilenode, uint64 stripe, TupleDesc tupleDescri
 	uint32 columnCount = tupleDescriptor->natts;
 	ScanKeyData scanKey[2];
 
-	uint64 storageId = LookupStorageId(relfilenode);
+	uint64 storageId = LookupStorageId(relfilelocator);
 
 	Oid columnarChunkOid = ColumnarChunkRelationId();
 	Relation columnarChunk = table_open(columnarChunkOid, AccessShareLock);
-	Relation index = index_open(ColumnarChunkIndexRelationId(), AccessShareLock);
 
 	ScanKeyInit(&scanKey[0], Anum_columnar_chunk_storageid,
-				BTEqualStrategyNumber, F_OIDEQ, UInt64GetDatum(storageId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(storageId));
 	ScanKeyInit(&scanKey[1], Anum_columnar_chunk_stripe,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(stripe));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(stripe));
 
-	SysScanDesc scanDescriptor = systable_beginscan_ordered(columnarChunk, index,
-															snapshot, 2, scanKey);
+	Oid indexId = ColumnarChunkIndexRelationId();
+	bool indexOk = OidIsValid(indexId);
+	SysScanDesc scanDescriptor = systable_beginscan(columnarChunk, indexId,
+													indexOk, snapshot, 2, scanKey);
+
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
+	{
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING, "chunk_pkey")));
+		loggedSlowMetadataAccessWarning = true;
+	}
 
 	StripeSkipList *chunkList = palloc0(sizeof(StripeSkipList));
 	chunkList->chunkCount = chunkCount;
@@ -721,8 +746,7 @@ ReadStripeSkipList(RelFileNode relfilenode, uint64 stripe, TupleDesc tupleDescri
 			palloc0(chunkCount * sizeof(ColumnChunkSkipNode));
 	}
 
-	while (HeapTupleIsValid(heapTuple = systable_getnext_ordered(scanDescriptor,
-																 ForwardScanDirection)))
+	while (HeapTupleIsValid(heapTuple = systable_getnext(scanDescriptor)))
 	{
 		Datum datumArray[Natts_columnar_chunk];
 		bool isNullArray[Natts_columnar_chunk];
@@ -787,8 +811,7 @@ ReadStripeSkipList(RelFileNode relfilenode, uint64 stripe, TupleDesc tupleDescri
 		}
 	}
 
-	systable_endscan_ordered(scanDescriptor);
-	index_close(index, AccessShareLock);
+	systable_endscan(scanDescriptor);
 	table_close(columnarChunk, AccessShareLock);
 
 	chunkList->chunkGroupRowCounts =
@@ -799,9 +822,9 @@ ReadStripeSkipList(RelFileNode relfilenode, uint64 stripe, TupleDesc tupleDescri
 
 
 /*
- * FindStripeByRowNumber returns StripeMetadata for the stripe whose
- * firstRowNumber is greater than given rowNumber. If no such stripe
- * exists, then returns NULL.
+ * FindStripeByRowNumber returns StripeMetadata for the stripe that has the
+ * smallest firstRowNumber among the stripes whose firstRowNumber is grater
+ * than given rowNumber. If no such stripe exists, then returns NULL.
  */
 StripeMetadata *
 FindNextStripeByRowNumber(Relation relation, uint64 rowNumber, Snapshot snapshot)
@@ -891,8 +914,7 @@ StripeGetHighestRowNumber(StripeMetadata *stripeMetadata)
 /*
  * StripeMetadataLookupRowNumber returns StripeMetadata for the stripe whose
  * firstRowNumber is less than or equal to (FIND_LESS_OR_EQUAL), or is
- * greater than (FIND_GREATER) given rowNumber by doing backward index
- * scan on stripe_first_row_number_idx.
+ * greater than (FIND_GREATER) given rowNumber.
  * If no such stripe exists, then returns NULL.
  */
 static StripeMetadata *
@@ -906,7 +928,7 @@ StripeMetadataLookupRowNumber(Relation relation, uint64 rowNumber, Snapshot snap
 	uint64 storageId = ColumnarStorageGetStorageId(relation, false);
 	ScanKeyData scanKey[2];
 	ScanKeyInit(&scanKey[0], Anum_columnar_stripe_storageid,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(storageId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(storageId));
 
 	StrategyNumber strategyNumber = InvalidStrategy;
 	RegProcedure procedure = InvalidOid;
@@ -921,33 +943,73 @@ StripeMetadataLookupRowNumber(Relation relation, uint64 rowNumber, Snapshot snap
 		procedure = F_INT8GT;
 	}
 	ScanKeyInit(&scanKey[1], Anum_columnar_stripe_first_row_number,
-				strategyNumber, procedure, UInt64GetDatum(rowNumber));
-
+				strategyNumber, procedure, Int64GetDatum(rowNumber));
 
 	Relation columnarStripes = table_open(ColumnarStripeRelationId(), AccessShareLock);
-	Relation index = index_open(ColumnarStripeFirstRowNumberIndexRelationId(),
-								AccessShareLock);
-	SysScanDesc scanDescriptor = systable_beginscan_ordered(columnarStripes, index,
-															snapshot, 2,
-															scanKey);
 
-	ScanDirection scanDirection = NoMovementScanDirection;
-	if (lookupMode == FIND_LESS_OR_EQUAL)
+	Oid indexId = ColumnarStripeFirstRowNumberIndexRelationId();
+	bool indexOk = OidIsValid(indexId);
+	SysScanDesc scanDescriptor = systable_beginscan(columnarStripes, indexId, indexOk,
+													snapshot, 2, scanKey);
+
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
 	{
-		scanDirection = BackwardScanDirection;
-	}
-	else if (lookupMode == FIND_GREATER)
-	{
-		scanDirection = ForwardScanDirection;
-	}
-	HeapTuple heapTuple = systable_getnext_ordered(scanDescriptor, scanDirection);
-	if (HeapTupleIsValid(heapTuple))
-	{
-		foundStripeMetadata = BuildStripeMetadata(columnarStripes, heapTuple);
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING,
+								 "stripe_first_row_number_idx")));
+		loggedSlowMetadataAccessWarning = true;
 	}
 
-	systable_endscan_ordered(scanDescriptor);
-	index_close(index, AccessShareLock);
+	if (indexOk)
+	{
+		ScanDirection scanDirection = NoMovementScanDirection;
+		if (lookupMode == FIND_LESS_OR_EQUAL)
+		{
+			scanDirection = BackwardScanDirection;
+		}
+		else if (lookupMode == FIND_GREATER)
+		{
+			scanDirection = ForwardScanDirection;
+		}
+		HeapTuple heapTuple = systable_getnext_ordered(scanDescriptor, scanDirection);
+		if (HeapTupleIsValid(heapTuple))
+		{
+			foundStripeMetadata = BuildStripeMetadata(columnarStripes, heapTuple);
+		}
+	}
+	else
+	{
+		HeapTuple heapTuple = NULL;
+		while (HeapTupleIsValid(heapTuple = systable_getnext(scanDescriptor)))
+		{
+			StripeMetadata *stripe = BuildStripeMetadata(columnarStripes, heapTuple);
+			if (!foundStripeMetadata)
+			{
+				/* first match */
+				foundStripeMetadata = stripe;
+			}
+			else if (lookupMode == FIND_LESS_OR_EQUAL &&
+					 stripe->firstRowNumber > foundStripeMetadata->firstRowNumber)
+			{
+				/*
+				 * Among the stripes with firstRowNumber less-than-or-equal-to given,
+				 * we're looking for the one with the greatest firstRowNumber.
+				 */
+				foundStripeMetadata = stripe;
+			}
+			else if (lookupMode == FIND_GREATER &&
+					 stripe->firstRowNumber < foundStripeMetadata->firstRowNumber)
+			{
+				/*
+				 * Among the stripes with firstRowNumber greater-than given,
+				 * we're looking for the one with the smallest firstRowNumber.
+				 */
+				foundStripeMetadata = stripe;
+			}
+		}
+	}
+
+	systable_endscan(scanDescriptor);
 	table_close(columnarStripes, AccessShareLock);
 
 	return foundStripeMetadata;
@@ -1021,8 +1083,8 @@ CheckStripeMetadataConsistency(StripeMetadata *stripeMetadata)
 
 /*
  * FindStripeWithHighestRowNumber returns StripeMetadata for the stripe that
- * has the row with highest rowNumber by doing backward index scan on
- * stripe_first_row_number_idx. If given relation is empty, then returns NULL.
+ * has the row with highest rowNumber. If given relation is empty, then returns
+ * NULL.
  */
 StripeMetadata *
 FindStripeWithHighestRowNumber(Relation relation, Snapshot snapshot)
@@ -1032,22 +1094,49 @@ FindStripeWithHighestRowNumber(Relation relation, Snapshot snapshot)
 	uint64 storageId = ColumnarStorageGetStorageId(relation, false);
 	ScanKeyData scanKey[1];
 	ScanKeyInit(&scanKey[0], Anum_columnar_stripe_storageid,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(storageId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(storageId));
 
 	Relation columnarStripes = table_open(ColumnarStripeRelationId(), AccessShareLock);
-	Relation index = index_open(ColumnarStripeFirstRowNumberIndexRelationId(),
-								AccessShareLock);
-	SysScanDesc scanDescriptor = systable_beginscan_ordered(columnarStripes, index,
-															snapshot, 1, scanKey);
 
-	HeapTuple heapTuple = systable_getnext_ordered(scanDescriptor, BackwardScanDirection);
-	if (HeapTupleIsValid(heapTuple))
+	Oid indexId = ColumnarStripeFirstRowNumberIndexRelationId();
+	bool indexOk = OidIsValid(indexId);
+	SysScanDesc scanDescriptor = systable_beginscan(columnarStripes, indexId, indexOk,
+													snapshot, 1, scanKey);
+
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
 	{
-		stripeWithHighestRowNumber = BuildStripeMetadata(columnarStripes, heapTuple);
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING,
+								 "stripe_first_row_number_idx")));
+		loggedSlowMetadataAccessWarning = true;
 	}
 
-	systable_endscan_ordered(scanDescriptor);
-	index_close(index, AccessShareLock);
+	if (indexOk)
+	{
+		/* do one-time fetch using the index */
+		HeapTuple heapTuple = systable_getnext_ordered(scanDescriptor,
+													   BackwardScanDirection);
+		if (HeapTupleIsValid(heapTuple))
+		{
+			stripeWithHighestRowNumber = BuildStripeMetadata(columnarStripes, heapTuple);
+		}
+	}
+	else
+	{
+		HeapTuple heapTuple = NULL;
+		while (HeapTupleIsValid(heapTuple = systable_getnext(scanDescriptor)))
+		{
+			StripeMetadata *stripe = BuildStripeMetadata(columnarStripes, heapTuple);
+			if (!stripeWithHighestRowNumber ||
+				stripe->firstRowNumber > stripeWithHighestRowNumber->firstRowNumber)
+			{
+				/* first or a greater match */
+				stripeWithHighestRowNumber = stripe;
+			}
+		}
+	}
+
+	systable_endscan(scanDescriptor);
 	table_close(columnarStripes, AccessShareLock);
 
 	return stripeWithHighestRowNumber;
@@ -1064,23 +1153,29 @@ ReadChunkGroupRowCounts(uint64 storageId, uint64 stripe, uint32 chunkGroupCount,
 {
 	Oid columnarChunkGroupOid = ColumnarChunkGroupRelationId();
 	Relation columnarChunkGroup = table_open(columnarChunkGroupOid, AccessShareLock);
-	Relation index = index_open(ColumnarChunkGroupIndexRelationId(), AccessShareLock);
 
 	ScanKeyData scanKey[2];
 	ScanKeyInit(&scanKey[0], Anum_columnar_chunkgroup_storageid,
-				BTEqualStrategyNumber, F_OIDEQ, UInt64GetDatum(storageId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(storageId));
 	ScanKeyInit(&scanKey[1], Anum_columnar_chunkgroup_stripe,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(stripe));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(stripe));
 
+	Oid indexId = ColumnarChunkGroupIndexRelationId();
+	bool indexOk = OidIsValid(indexId);
 	SysScanDesc scanDescriptor =
-		systable_beginscan_ordered(columnarChunkGroup, index, snapshot, 2, scanKey);
+		systable_beginscan(columnarChunkGroup, indexId, indexOk, snapshot, 2, scanKey);
+
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
+	{
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING, "chunk_group_pkey")));
+		loggedSlowMetadataAccessWarning = true;
+	}
 
-	uint32 chunkGroupIndex = 0;
 	HeapTuple heapTuple = NULL;
 	uint32 *chunkGroupRowCounts = palloc0(chunkGroupCount * sizeof(uint32));
 
-	while (HeapTupleIsValid(heapTuple = systable_getnext_ordered(scanDescriptor,
-																 ForwardScanDirection)))
+	while (HeapTupleIsValid(heapTuple = systable_getnext(scanDescriptor)))
 	{
 		Datum datumArray[Natts_columnar_chunkgroup];
 		bool isNullArray[Natts_columnar_chunkgroup];
@@ -1091,24 +1186,16 @@ ReadChunkGroupRowCounts(uint64 storageId, uint64 stripe, uint32 chunkGroupCount,
 
 		uint32 tupleChunkGroupIndex =
 			DatumGetUInt32(datumArray[Anum_columnar_chunkgroup_chunk - 1]);
-		if (chunkGroupIndex >= chunkGroupCount ||
-			tupleChunkGroupIndex != chunkGroupIndex)
+		if (tupleChunkGroupIndex >= chunkGroupCount)
 		{
 			elog(ERROR, "unexpected chunk group");
 		}
 
-		chunkGroupRowCounts[chunkGroupIndex] =
+		chunkGroupRowCounts[tupleChunkGroupIndex] =
 			(uint32) DatumGetUInt64(datumArray[Anum_columnar_chunkgroup_row_count - 1]);
-		chunkGroupIndex++;
 	}
 
-	if (chunkGroupIndex != chunkGroupCount)
-	{
-		elog(ERROR, "unexpected chunk group count");
-	}
-
-	systable_endscan_ordered(scanDescriptor);
-	index_close(index, AccessShareLock);
+	systable_endscan(scanDescriptor);
 	table_close(columnarChunkGroup, AccessShareLock);
 
 	return chunkGroupRowCounts;
@@ -1161,13 +1248,13 @@ InsertEmptyStripeMetadataRow(uint64 storageId, uint64 stripeId, uint32 columnCou
 
 
 /*
- * StripesForRelfilenode returns a list of StripeMetadata for stripes
+ * StripesForRelfilelocator returns a list of StripeMetadata for stripes
  * of the given relfilenode.
  */
 List *
-StripesForRelfilenode(RelFileNode relfilenode)
+StripesForRelfilelocator(RelFileLocator relfilelocator)
 {
-	uint64 storageId = LookupStorageId(relfilenode);
+	uint64 storageId = LookupStorageId(relfilelocator);
 
 	return ReadDataFileStripeList(storageId, GetTransactionSnapshot());
 }
@@ -1182,9 +1269,9 @@ StripesForRelfilenode(RelFileNode relfilenode)
  * returns 0.
  */
 uint64
-GetHighestUsedAddress(RelFileNode relfilenode)
+GetHighestUsedAddress(RelFileLocator relfilelocator)
 {
-	uint64 storageId = LookupStorageId(relfilenode);
+	uint64 storageId = LookupStorageId(relfilelocator);
 
 	uint64 highestUsedAddress = 0;
 	uint64 highestUsedId = 0;
@@ -1298,21 +1385,27 @@ UpdateStripeMetadataRow(uint64 storageId, uint64 stripeId, bool *update,
 
 	ScanKeyData scanKey[2];
 	ScanKeyInit(&scanKey[0], Anum_columnar_stripe_storageid,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(storageId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(storageId));
 	ScanKeyInit(&scanKey[1], Anum_columnar_stripe_stripe,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(stripeId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(stripeId));
 
 	Oid columnarStripesOid = ColumnarStripeRelationId();
 
 	Relation columnarStripes = table_open(columnarStripesOid, AccessShareLock);
-	Relation columnarStripePkeyIndex = index_open(ColumnarStripePKeyIndexRelationId(),
-												  AccessShareLock);
 
-	SysScanDesc scanDescriptor = systable_beginscan_ordered(columnarStripes,
-															columnarStripePkeyIndex,
-															&dirtySnapshot, 2, scanKey);
+	Oid indexId = ColumnarStripePKeyIndexRelationId();
+	bool indexOk = OidIsValid(indexId);
+	SysScanDesc scanDescriptor = systable_beginscan(columnarStripes, indexId, indexOk,
+													&dirtySnapshot, 2, scanKey);
 
-	HeapTuple oldTuple = systable_getnext_ordered(scanDescriptor, ForwardScanDirection);
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
+	{
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING, "stripe_pkey")));
+		loggedSlowMetadataAccessWarning = true;
+	}
+
+	HeapTuple oldTuple = systable_getnext(scanDescriptor);
 	if (!HeapTupleIsValid(oldTuple))
 	{
 		ereport(ERROR, (errmsg("attempted to modify an unexpected stripe, "
@@ -1347,8 +1440,7 @@ UpdateStripeMetadataRow(uint64 storageId, uint64 stripeId, bool *update,
 
 	CommandCounterIncrement();
 
-	systable_endscan_ordered(scanDescriptor);
-	index_close(columnarStripePkeyIndex, AccessShareLock);
+	systable_endscan(scanDescriptor);
 	table_close(columnarStripes, AccessShareLock);
 
 	/* return StripeMetadata object built from modified tuple */
@@ -1359,6 +1451,10 @@ UpdateStripeMetadataRow(uint64 storageId, uint64 stripeId, bool *update,
 /*
  * ReadDataFileStripeList reads the stripe list for a given storageId
  * in the given snapshot.
+ *
+ * Doesn't sort the stripes by their ids before returning if
+ * stripe_first_row_number_idx is not available --normally can only happen
+ * during pg upgrades.
  */
 static List *
 ReadDataFileStripeList(uint64 storageId, Snapshot snapshot)
@@ -1368,27 +1464,32 @@ ReadDataFileStripeList(uint64 storageId, Snapshot snapshot)
 	HeapTuple heapTuple;
 
 	ScanKeyInit(&scanKey[0], Anum_columnar_stripe_storageid,
-				BTEqualStrategyNumber, F_OIDEQ, Int32GetDatum(storageId));
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(storageId));
 
 	Oid columnarStripesOid = ColumnarStripeRelationId();
 
 	Relation columnarStripes = table_open(columnarStripesOid, AccessShareLock);
-	Relation index = index_open(ColumnarStripeFirstRowNumberIndexRelationId(),
-								AccessShareLock);
 
-	SysScanDesc scanDescriptor = systable_beginscan_ordered(columnarStripes, index,
-															snapshot, 1,
-															scanKey);
+	Oid indexId = ColumnarStripeFirstRowNumberIndexRelationId();
+	bool indexOk = OidIsValid(indexId);
+	SysScanDesc scanDescriptor = systable_beginscan(columnarStripes, indexId,
+													indexOk, snapshot, 1, scanKey);
 
-	while (HeapTupleIsValid(heapTuple = systable_getnext_ordered(scanDescriptor,
-																 ForwardScanDirection)))
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
+	{
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING,
+								 "stripe_first_row_number_idx")));
+		loggedSlowMetadataAccessWarning = true;
+	}
+
+	while (HeapTupleIsValid(heapTuple = systable_getnext(scanDescriptor)))
 	{
 		StripeMetadata *stripeMetadata = BuildStripeMetadata(columnarStripes, heapTuple);
 		stripeMetadataList = lappend(stripeMetadataList, stripeMetadata);
 	}
 
-	systable_endscan_ordered(scanDescriptor);
-	index_close(index, AccessShareLock);
+	systable_endscan(scanDescriptor);
 	table_close(columnarStripes, AccessShareLock);
 
 	return stripeMetadataList;
@@ -1451,7 +1552,7 @@ BuildStripeMetadata(Relation columnarStripes, HeapTuple heapTuple)
  * metadata tables.
  */
 void
-DeleteMetadataRows(RelFileNode relfilenode)
+DeleteMetadataRows(RelFileLocator relfilelocator)
 {
 	/*
 	 * During a restore for binary upgrade, metadata tables and indexes may or
@@ -1462,7 +1563,7 @@ DeleteMetadataRows(RelFileNode relfilenode)
 		return;
 	}
 
-	uint64 storageId = LookupStorageId(relfilenode);
+	uint64 storageId = LookupStorageId(relfilelocator);
 
 	DeleteStorageFromColumnarMetadataTable(ColumnarStripeRelationId(),
 										   Anum_columnar_stripe_storageid,
@@ -1490,7 +1591,7 @@ DeleteStorageFromColumnarMetadataTable(Oid metadataTableId,
 {
 	ScanKeyData scanKey[1];
 	ScanKeyInit(&scanKey[0], storageIdAtrrNumber, BTEqualStrategyNumber,
-				F_INT8EQ, UInt64GetDatum(storageId));
+				F_INT8EQ, Int64GetDatum(storageId));
 
 	Relation metadataTable = try_relation_open(metadataTableId, AccessShareLock);
 	if (metadataTable == NULL)
@@ -1499,25 +1600,30 @@ DeleteStorageFromColumnarMetadataTable(Oid metadataTableId,
 		return;
 	}
 
-	Relation index = index_open(storageIdIndexId, AccessShareLock);
+	bool indexOk = OidIsValid(storageIdIndexId);
+	SysScanDesc scanDescriptor = systable_beginscan(metadataTable, storageIdIndexId,
+													indexOk, NULL, 1, scanKey);
 
-	SysScanDesc scanDescriptor = systable_beginscan_ordered(metadataTable, index, NULL,
-															1, scanKey);
+	static bool loggedSlowMetadataAccessWarning = false;
+	if (!indexOk && !loggedSlowMetadataAccessWarning)
+	{
+		ereport(WARNING, (errmsg(SLOW_METADATA_ACCESS_WARNING,
+								 "on a columnar metadata table")));
+		loggedSlowMetadataAccessWarning = true;
+	}
 
 	ModifyState *modifyState = StartModifyRelation(metadataTable);
 
 	HeapTuple heapTuple;
-	while (HeapTupleIsValid(heapTuple = systable_getnext_ordered(scanDescriptor,
-																 ForwardScanDirection)))
+	while (HeapTupleIsValid(heapTuple = systable_getnext(scanDescriptor)))
 	{
 		DeleteTupleAndEnforceConstraints(modifyState, heapTuple);
 	}
 
-	systable_endscan_ordered(scanDescriptor);
+	systable_endscan(scanDescriptor);
 
 	FinishModifyRelation(modifyState);
 
-	index_close(index, AccessShareLock);
 	table_close(metadataTable, AccessShareLock);
 }
 
@@ -1530,12 +1636,8 @@ StartModifyRelation(Relation rel)
 {
 	EState *estate = create_estate_for_relation(rel);
 
-#if PG_VERSION_NUM >= PG_VERSION_14
 	ResultRelInfo *resultRelInfo = makeNode(ResultRelInfo);
 	InitResultRelInfo(resultRelInfo, rel, 1, NULL, 0);
-#else
-	ResultRelInfo *resultRelInfo = estate->es_result_relation_info;
-#endif
 
 	/* ExecSimpleRelationInsert, ... require caller to open indexes */
 	ExecOpenIndices(resultRelInfo, false);
@@ -1565,7 +1667,7 @@ InsertTupleAndEnforceConstraints(ModifyState *state, Datum *values, bool *nulls)
 	ExecStoreHeapTuple(tuple, slot, false);
 
 	/* use ExecSimpleRelationInsert to enforce constraints */
-	ExecSimpleRelationInsert_compat(state->resultRelInfo, state->estate, slot);
+	ExecSimpleRelationInsert(state->resultRelInfo, state->estate, slot);
 }
 
 
@@ -1583,7 +1685,7 @@ DeleteTupleAndEnforceConstraints(ModifyState *state, HeapTuple heapTuple)
 	simple_heap_delete(state->rel, tid);
 
 	/* execute AFTER ROW DELETE Triggers to enforce constraints */
-	ExecARDeleteTriggers_compat(estate, resultRelInfo, tid, NULL, NULL, false);
+	ExecARDeleteTriggers(estate, resultRelInfo, tid, NULL, NULL, false);
 }
 
 
@@ -1596,12 +1698,8 @@ FinishModifyRelation(ModifyState *state)
 	ExecCloseIndices(state->resultRelInfo);
 
 	AfterTriggerEndQuery(state->estate);
-#if PG_VERSION_NUM >= PG_VERSION_14
 	ExecCloseResultRelations(state->estate);
 	ExecCloseRangeTableRelations(state->estate);
-#else
-	ExecCleanUpTriggerState(state->estate);
-#endif
 	ExecResetTupleTable(state->estate->es_tupleTable, false);
 	FreeExecutorState(state->estate);
 
@@ -1628,15 +1726,13 @@ create_estate_for_relation(Relation rel)
 	rte->relid = RelationGetRelid(rel);
 	rte->relkind = rel->rd_rel->relkind;
 	rte->rellockmode = AccessShareLock;
+
+#if PG_VERSION_NUM >= PG_VERSION_16
+	List *perminfos = NIL;
+	addRTEPermissionInfo(&perminfos, rte);
+	ExecInitRangeTable(estate, list_make1(rte), perminfos);
+#else
 	ExecInitRangeTable(estate, list_make1(rte));
-
-#if PG_VERSION_NUM < PG_VERSION_14
-	ResultRelInfo *resultRelInfo = makeNode(ResultRelInfo);
-	InitResultRelInfo(resultRelInfo, rel, 1, NULL, 0);
-
-	estate->es_result_relations = resultRelInfo;
-	estate->es_num_result_relations = 1;
-	estate->es_result_relation_info = resultRelInfo;
 #endif
 
 	estate->es_output_cid = GetCurrentCommandId(true);
@@ -1650,6 +1746,9 @@ create_estate_for_relation(Relation rel)
 
 /*
  * DatumToBytea serializes a datum into a bytea value.
+ *
+ * Since we don't want to limit datum size to RSIZE_MAX unnecessarily,
+ * we use memcpy instead of memcpy_s several places in this function.
  */
 static bytea *
 DatumToBytea(Datum value, Form_pg_attribute attrForm)
@@ -1666,19 +1765,16 @@ DatumToBytea(Datum value, Form_pg_attribute attrForm)
 			Datum tmp;
 			store_att_byval(&tmp, value, attrForm->attlen);
 
-			memcpy_s(VARDATA(result), datumLength + VARHDRSZ,
-					 &tmp, attrForm->attlen);
+			memcpy(VARDATA(result), &tmp, attrForm->attlen); /* IGNORE-BANNED */
 		}
 		else
 		{
-			memcpy_s(VARDATA(result), datumLength + VARHDRSZ,
-					 DatumGetPointer(value), attrForm->attlen);
+			memcpy(VARDATA(result), DatumGetPointer(value), attrForm->attlen); /* IGNORE-BANNED */
 		}
 	}
 	else
 	{
-		memcpy_s(VARDATA(result), datumLength + VARHDRSZ,
-				 DatumGetPointer(value), datumLength);
+		memcpy(VARDATA(result), DatumGetPointer(value), datumLength); /* IGNORE-BANNED */
 	}
 
 	return result;
@@ -1697,8 +1793,12 @@ ByteaToDatum(bytea *bytes, Form_pg_attribute attrForm)
 	 * after the byteaDatum is freed.
 	 */
 	char *binaryDataCopy = palloc0(VARSIZE_ANY_EXHDR(bytes));
-	memcpy_s(binaryDataCopy, VARSIZE_ANY_EXHDR(bytes),
-			 VARDATA_ANY(bytes), VARSIZE_ANY_EXHDR(bytes));
+
+	/*
+	 * We use IGNORE-BANNED here since we don't want to limit datum size to
+	 * RSIZE_MAX unnecessarily.
+	 */
+	memcpy(binaryDataCopy, VARDATA_ANY(bytes), VARSIZE_ANY_EXHDR(bytes)); /* IGNORE-BANNED */
 
 	return fetch_att(binaryDataCopy, attrForm->attbyval, attrForm->attlen);
 }
@@ -1837,10 +1937,11 @@ ColumnarNamespaceId(void)
  * false if the relation doesn't have a meta page yet.
  */
 static uint64
-LookupStorageId(RelFileNode relfilenode)
+LookupStorageId(RelFileLocator relfilelocator)
 {
-	Oid relationId = RelidByRelfilenode(relfilenode.spcNode,
-										relfilenode.relNode);
+	Oid relationId = RelidByRelfilenumber(RelationTablespace_compat(relfilelocator),
+										  RelationPhysicalIdentifierNumber_compat(
+											  relfilelocator));
 
 	Relation relation = relation_open(relationId, AccessShareLock);
 	uint64 storageId = ColumnarStorageGetStorageId(relation, false);
@@ -1871,7 +1972,7 @@ columnar_relation_storageid(PG_FUNCTION_ARGS)
 	Oid relationId = PG_GETARG_OID(0);
 	Relation relation = relation_open(relationId, AccessShareLock);
 
-	if (!pg_class_ownercheck(relationId, GetUserId()))
+	if (!object_ownercheck(RelationRelationId, relationId, GetUserId()))
 	{
 		aclcheck_error(ACLCHECK_NOT_OWNER, OBJECT_TABLE,
 					   get_rel_name(relationId));
@@ -1940,7 +2041,7 @@ GetHighestUsedRowNumber(uint64 storageId)
 	List *stripeMetadataList = ReadDataFileStripeList(storageId,
 													  GetTransactionSnapshot());
 	StripeMetadata *stripeMetadata = NULL;
-	foreach_ptr(stripeMetadata, stripeMetadataList)
+	foreach_declared_ptr(stripeMetadata, stripeMetadataList)
 	{
 		highestRowNumber = Max(highestRowNumber,
 							   StripeGetHighestRowNumber(stripeMetadata));

src/backend/columnar/columnar_reader.c

@@ -22,16 +22,15 @@
 #include "access/xact.h"
 #include "catalog/pg_am.h"
 #include "commands/defrem.h"
-#include "distributed/listutils.h"
 #include "nodes/makefuncs.h"
 #include "nodes/nodeFuncs.h"
-#include "optimizer/optimizer.h"
 #include "optimizer/clauses.h"
+#include "optimizer/optimizer.h"
 #include "optimizer/restrictinfo.h"
 #include "storage/fd.h"
 #include "utils/guc.h"
-#include "utils/memutils.h"
 #include "utils/lsyscache.h"
+#include "utils/memutils.h"
 #include "utils/rel.h"
 
 #include "columnar/columnar.h"
@@ -39,6 +38,8 @@
 #include "columnar/columnar_tableam.h"
 #include "columnar/columnar_version_compat.h"
 
+#include "distributed/listutils.h"
+
 #define UNEXPECTED_STRIPE_READ_ERR_MSG \
 	"attempted to read an unexpected stripe while reading columnar " \
 	"table %s, stripe with id=" UINT64_FORMAT " is not flushed"
@@ -254,8 +255,9 @@ ColumnarReadFlushPendingWrites(ColumnarReadState *readState)
 {
 	Assert(!readState->snapshotRegisteredByUs);
 
-	Oid relfilenode = readState->relation->rd_node.relNode;
-	FlushWriteStateForRelfilenode(relfilenode, GetCurrentSubTransactionId());
+	RelFileNumber relfilenumber = RelationPhysicalIdentifierNumber_compat(
+		RelationPhysicalIdentifier_compat(readState->relation));
+	FlushWriteStateForRelfilenumber(relfilenumber, GetCurrentSubTransactionId());
 
 	if (readState->snapshot == InvalidSnapshot || !IsMVCCSnapshot(readState->snapshot))
 	{
@@ -878,7 +880,7 @@ ReadChunkGroupNextRow(ChunkGroupReadState *chunkGroupReadState, Datum *columnVal
 	memset(columnNulls, true, sizeof(bool) * chunkGroupReadState->columnCount);
 
 	int attno;
-	foreach_int(attno, chunkGroupReadState->projectedColumnList)
+	foreach_declared_int(attno, chunkGroupReadState->projectedColumnList)
 	{
 		const ChunkData *chunkGroupData = chunkGroupReadState->chunkGroupData;
 		const int rowIndex = chunkGroupReadState->currentRow;
@@ -984,7 +986,8 @@ ColumnarTableRowCount(Relation relation)
 {
 	ListCell *stripeMetadataCell = NULL;
 	uint64 totalRowCount = 0;
-	List *stripeList = StripesForRelfilenode(relation->rd_node);
+	List *stripeList = StripesForRelfilelocator(RelationPhysicalIdentifier_compat(
+													relation));
 
 	foreach(stripeMetadataCell, stripeList)
 	{
@@ -1012,7 +1015,8 @@ LoadFilteredStripeBuffers(Relation relation, StripeMetadata *stripeMetadata,
 
 	bool *projectedColumnMask = ProjectedColumnMask(columnCount, projectedColumnList);
 
-	StripeSkipList *stripeSkipList = ReadStripeSkipList(relation->rd_node,
+	StripeSkipList *stripeSkipList = ReadStripeSkipList(RelationPhysicalIdentifier_compat(
+															relation),
 														stripeMetadata->id,
 														tupleDescriptor,
 														stripeMetadata->chunkCount,
@@ -1485,7 +1489,7 @@ ProjectedColumnMask(uint32 columnCount, List *projectedColumnList)
 	bool *projectedColumnMask = palloc0(columnCount * sizeof(bool));
 	int attno;
 
-	foreach_int(attno, projectedColumnList)
+	foreach_declared_int(attno, projectedColumnList)
 	{
 		/* attno is 1-indexed; projectedColumnMask is 0-indexed */
 		int columnIndex = attno - 1;
@@ -1557,7 +1561,7 @@ DeserializeDatumArray(StringInfo datumBuffer, bool *existsArray, uint32 datumCou
 										   datumTypeLength);
 		currentDatumDataOffset = att_addlength_datum(currentDatumDataOffset,
 													 datumTypeLength,
-													 currentDatumDataPointer);
+													 datumArray[datumIndex]);
 		currentDatumDataOffset = att_align_nominal(currentDatumDataOffset,
 												   datumTypeAlign);
 

src/backend/columnar/columnar_storage.c

@@ -36,11 +36,11 @@
 
 #include "postgres.h"
 
+#include "miscadmin.h"
 #include "safe_lib.h"
 
 #include "access/generic_xlog.h"
 #include "catalog/storage.h"
-#include "miscadmin.h"
 #include "storage/bufmgr.h"
 #include "storage/lmgr.h"
 
@@ -169,7 +169,11 @@ ColumnarStorageInit(SMgrRelation srel, uint64 storageId)
 	}
 
 	/* create two pages */
+#if PG_VERSION_NUM >= PG_VERSION_16
+	PGIOAlignedBlock block;
+#else
 	PGAlignedBlock block;
+#endif
 	Page page = block.data;
 
 	/* write metapage */
@@ -188,7 +192,7 @@ ColumnarStorageInit(SMgrRelation srel, uint64 storageId)
 			 (char *) &metapage, sizeof(ColumnarMetapage));
 	phdr->pd_lower += sizeof(ColumnarMetapage);
 
-	log_newpage(&srel->smgr_rnode.node, MAIN_FORKNUM,
+	log_newpage(RelationPhysicalIdentifierBackend_compat(&srel), MAIN_FORKNUM,
 				COLUMNAR_METAPAGE_BLOCKNO, page, true);
 	PageSetChecksumInplace(page, COLUMNAR_METAPAGE_BLOCKNO);
 	smgrextend(srel, MAIN_FORKNUM, COLUMNAR_METAPAGE_BLOCKNO, page, true);
@@ -196,7 +200,7 @@ ColumnarStorageInit(SMgrRelation srel, uint64 storageId)
 	/* write empty page */
 	PageInit(page, BLCKSZ, 0);
 
-	log_newpage(&srel->smgr_rnode.node, MAIN_FORKNUM,
+	log_newpage(RelationPhysicalIdentifierBackend_compat(&srel), MAIN_FORKNUM,
 				COLUMNAR_EMPTY_BLOCKNO, page, true);
 	PageSetChecksumInplace(page, COLUMNAR_EMPTY_BLOCKNO);
 	smgrextend(srel, MAIN_FORKNUM, COLUMNAR_EMPTY_BLOCKNO, page, true);

src/backend/columnar/columnar_tableam.c

@@ -1,41 +1,38 @@
-#include "citus_version.h"
+#include <math.h>
 
 #include "postgres.h"
 
-#include <math.h>
-
 #include "miscadmin.h"
+#include "pgstat.h"
+#include "safe_lib.h"
 
+#include "access/detoast.h"
 #include "access/genam.h"
 #include "access/heapam.h"
 #include "access/multixact.h"
 #include "access/rewriteheap.h"
 #include "access/tableam.h"
 #include "access/tsmapi.h"
-#include "access/detoast.h"
 #include "access/xact.h"
 #include "catalog/catalog.h"
 #include "catalog/index.h"
 #include "catalog/namespace.h"
 #include "catalog/objectaccess.h"
 #include "catalog/pg_am.h"
+#include "catalog/pg_extension.h"
 #include "catalog/pg_publication.h"
 #include "catalog/pg_trigger.h"
-#include "catalog/pg_extension.h"
 #include "catalog/storage.h"
 #include "catalog/storage_xlog.h"
 #include "commands/defrem.h"
+#include "commands/extension.h"
 #include "commands/progress.h"
 #include "commands/vacuum.h"
-#include "commands/extension.h"
 #include "executor/executor.h"
 #include "nodes/makefuncs.h"
 #include "optimizer/plancat.h"
-#include "pgstat.h"
-#include "safe_lib.h"
 #include "storage/bufmgr.h"
 #include "storage/bufpage.h"
-#include "storage/bufmgr.h"
 #include "storage/lmgr.h"
 #include "storage/predicate.h"
 #include "storage/procarray.h"
@@ -43,17 +40,22 @@
 #include "tcop/utility.h"
 #include "utils/builtins.h"
 #include "utils/fmgroids.h"
+#include "utils/lsyscache.h"
 #include "utils/memutils.h"
 #include "utils/pg_rusage.h"
 #include "utils/rel.h"
 #include "utils/relcache.h"
-#include "utils/lsyscache.h"
 #include "utils/syscache.h"
+
+#include "citus_version.h"
+#include "pg_version_compat.h"
+
 #include "columnar/columnar.h"
 #include "columnar/columnar_customscan.h"
 #include "columnar/columnar_storage.h"
 #include "columnar/columnar_tableam.h"
 #include "columnar/columnar_version_compat.h"
+
 #include "distributed/listutils.h"
 
 /*
@@ -115,9 +117,7 @@ static RangeVar * ColumnarProcessAlterTable(AlterTableStmt *alterTableStmt,
 											List **columnarOptions);
 static void ColumnarProcessUtility(PlannedStmt *pstmt,
 								   const char *queryString,
-#if PG_VERSION_NUM >= PG_VERSION_14
 								   bool readOnlyTree,
-#endif
 								   ProcessUtilityContext context,
 								   ParamListInfo params,
 								   struct QueryEnvironment *queryEnv,
@@ -208,7 +208,8 @@ columnar_beginscan_extended(Relation relation, Snapshot snapshot,
 							uint32 flags, Bitmapset *attr_needed, List *scanQual)
 {
 	CheckCitusColumnarVersion(ERROR);
-	Oid relfilenode = relation->rd_node.relNode;
+	RelFileNumber relfilenumber = RelationPhysicalIdentifierNumber_compat(
+		RelationPhysicalIdentifier_compat(relation));
 
 	/*
 	 * A memory context to use for scan-wide data, including the lazily
@@ -238,7 +239,7 @@ columnar_beginscan_extended(Relation relation, Snapshot snapshot,
 	scan->scanQual = copyObject(scanQual);
 	scan->scanContext = scanContext;
 
-	if (PendingWritesInUpperTransactions(relfilenode, GetCurrentSubTransactionId()))
+	if (PendingWritesInUpperTransactions(relfilenumber, GetCurrentSubTransactionId()))
 	{
 		elog(ERROR,
 			 "cannot read from table when there is unflushed data in upper transactions");
@@ -434,8 +435,9 @@ columnar_index_fetch_begin(Relation rel)
 {
 	CheckCitusColumnarVersion(ERROR);
 
-	Oid relfilenode = rel->rd_node.relNode;
-	if (PendingWritesInUpperTransactions(relfilenode, GetCurrentSubTransactionId()))
+	RelFileNumber relfilenumber = RelationPhysicalIdentifierNumber_compat(
+		RelationPhysicalIdentifier_compat(rel));
+	if (PendingWritesInUpperTransactions(relfilenumber, GetCurrentSubTransactionId()))
 	{
 		/* XXX: maybe we can just flush the data and continue */
 		elog(ERROR, "cannot read from index when there is unflushed data in "
@@ -665,7 +667,6 @@ columnar_tuple_satisfies_snapshot(Relation rel, TupleTableSlot *slot,
 }
 
 
-#if PG_VERSION_NUM >= PG_VERSION_14
 static TransactionId
 columnar_index_delete_tuples(Relation rel,
 							 TM_IndexDeleteOp *delstate)
@@ -714,19 +715,6 @@ columnar_index_delete_tuples(Relation rel,
 }
 
 
-#else
-static TransactionId
-columnar_compute_xid_horizon_for_tuples(Relation rel,
-										ItemPointerData *tids,
-										int nitems)
-{
-	elog(ERROR, "columnar_compute_xid_horizon_for_tuples not implemented");
-}
-
-
-#endif
-
-
 static void
 columnar_tuple_insert(Relation relation, TupleTableSlot *slot, CommandId cid,
 					  int options, BulkInsertState bistate)
@@ -739,7 +727,9 @@ columnar_tuple_insert(Relation relation, TupleTableSlot *slot, CommandId cid,
 	 */
 	ColumnarWriteState *writeState = columnar_init_write_state(relation,
 															   RelationGetDescr(relation),
+															   slot->tts_tableOid,
 															   GetCurrentSubTransactionId());
+
 	MemoryContext oldContext = MemoryContextSwitchTo(ColumnarWritePerTupleContext(
 														 writeState));
 
@@ -781,8 +771,14 @@ columnar_multi_insert(Relation relation, TupleTableSlot **slots, int ntuples,
 {
 	CheckCitusColumnarVersion(ERROR);
 
+	/*
+	 * The callback to .multi_insert is table_multi_insert() and this is only used for the COPY
+	 * command, so slot[i]->tts_tableoid will always be equal to relation->id. Thus, we can send
+	 * RelationGetRelid(relation) as the tupSlotTableOid
+	 */
 	ColumnarWriteState *writeState = columnar_init_write_state(relation,
 															   RelationGetDescr(relation),
+															   RelationGetRelid(relation),
 															   GetCurrentSubTransactionId());
 
 	ColumnarCheckLogicalReplication(relation);
@@ -823,7 +819,7 @@ static TM_Result
 columnar_tuple_update(Relation relation, ItemPointer otid, TupleTableSlot *slot,
 					  CommandId cid, Snapshot snapshot, Snapshot crosscheck,
 					  bool wait, TM_FailureData *tmfd,
-					  LockTupleMode *lockmode, bool *update_indexes)
+					  LockTupleMode *lockmode, TU_UpdateIndexes *update_indexes)
 {
 	elog(ERROR, "columnar_tuple_update not implemented");
 }
@@ -849,11 +845,11 @@ columnar_finish_bulk_insert(Relation relation, int options)
 
 
 static void
-columnar_relation_set_new_filenode(Relation rel,
-								   const RelFileNode *newrnode,
-								   char persistence,
-								   TransactionId *freezeXid,
-								   MultiXactId *minmulti)
+columnar_relation_set_new_filelocator(Relation rel,
+									  const RelFileLocator *newrlocator,
+									  char persistence,
+									  TransactionId *freezeXid,
+									  MultiXactId *minmulti)
 {
 	CheckCitusColumnarVersion(ERROR);
 
@@ -869,16 +865,19 @@ columnar_relation_set_new_filenode(Relation rel,
 	 * state. If they are equal, this is a new relation object and we don't
 	 * need to clean anything.
 	 */
-	if (rel->rd_node.relNode != newrnode->relNode)
+	if (RelationPhysicalIdentifierNumber_compat(RelationPhysicalIdentifier_compat(rel)) !=
+		RelationPhysicalIdentifierNumberPtr_compat(newrlocator))
 	{
-		MarkRelfilenodeDropped(rel->rd_node.relNode, GetCurrentSubTransactionId());
+		MarkRelfilenumberDropped(RelationPhysicalIdentifierNumber_compat(
+									 RelationPhysicalIdentifier_compat(rel)),
+								 GetCurrentSubTransactionId());
 
-		DeleteMetadataRows(rel->rd_node);
+		DeleteMetadataRows(RelationPhysicalIdentifier_compat(rel));
 	}
 
 	*freezeXid = RecentXmin;
 	*minmulti = GetOldestMultiXactId();
-	SMgrRelation srel = RelationCreateStorage_compat(*newrnode, persistence, true);
+	SMgrRelation srel = RelationCreateStorage(*newrlocator, persistence, true);
 
 	ColumnarStorageInit(srel, ColumnarMetadataNewStorageId());
 	InitColumnarOptions(rel->rd_id);
@@ -893,12 +892,12 @@ static void
 columnar_relation_nontransactional_truncate(Relation rel)
 {
 	CheckCitusColumnarVersion(ERROR);
-	RelFileNode relfilenode = rel->rd_node;
+	RelFileLocator relfilelocator = RelationPhysicalIdentifier_compat(rel);
 
-	NonTransactionDropWriteState(relfilenode.relNode);
+	NonTransactionDropWriteState(RelationPhysicalIdentifierNumber_compat(relfilelocator));
 
 	/* Delete old relfilenode metadata */
-	DeleteMetadataRows(relfilenode);
+	DeleteMetadataRows(relfilelocator);
 
 	/*
 	 * No need to set new relfilenode, since the table was created in this
@@ -915,7 +914,7 @@ columnar_relation_nontransactional_truncate(Relation rel)
 
 
 static void
-columnar_relation_copy_data(Relation rel, const RelFileNode *newrnode)
+columnar_relation_copy_data(Relation rel, const RelFileLocator *newrnode)
 {
 	elog(ERROR, "columnar_relation_copy_data not implemented");
 }
@@ -961,7 +960,8 @@ columnar_relation_copy_for_cluster(Relation OldHeap, Relation NewHeap,
 	ColumnarOptions columnarOptions = { 0 };
 	ReadColumnarOptions(OldHeap->rd_id, &columnarOptions);
 
-	ColumnarWriteState *writeState = ColumnarBeginWrite(NewHeap->rd_node,
+	ColumnarWriteState *writeState = ColumnarBeginWrite(RelationPhysicalIdentifier_compat(
+															NewHeap),
 														columnarOptions,
 														targetDesc);
 
@@ -1036,7 +1036,8 @@ NeededColumnsList(TupleDesc tupdesc, Bitmapset *attr_needed)
 static uint64
 ColumnarTableTupleCount(Relation relation)
 {
-	List *stripeList = StripesForRelfilenode(relation->rd_node);
+	List *stripeList = StripesForRelfilelocator(RelationPhysicalIdentifier_compat(
+													relation));
 	uint64 tupleCount = 0;
 
 	ListCell *lc = NULL;
@@ -1099,12 +1100,38 @@ columnar_vacuum_rel(Relation rel, VacuumParams *params,
 	List *indexList = RelationGetIndexList(rel);
 	int nindexes = list_length(indexList);
 
+#if PG_VERSION_NUM >= PG_VERSION_16
+	struct VacuumCutoffs cutoffs;
+	vacuum_get_cutoffs(rel, params, &cutoffs);
+
+	Assert(MultiXactIdPrecedesOrEquals(cutoffs.MultiXactCutoff, cutoffs.OldestMxact));
+	Assert(TransactionIdPrecedesOrEquals(cutoffs.FreezeLimit, cutoffs.OldestXmin));
+
+	/*
+	 * Columnar storage doesn't hold any transaction IDs, so we can always
+	 * just advance to the most aggressive value.
+	 */
+	TransactionId newRelFrozenXid = cutoffs.OldestXmin;
+	MultiXactId newRelminMxid = cutoffs.OldestMxact;
+	double new_live_tuples = ColumnarTableTupleCount(rel);
+
+	/* all visible pages are always 0 */
+	BlockNumber new_rel_allvisible = 0;
+
+	bool frozenxid_updated;
+	bool minmulti_updated;
+
+	vac_update_relstats(rel, new_rel_pages, new_live_tuples,
+						new_rel_allvisible, nindexes > 0,
+						newRelFrozenXid, newRelminMxid,
+						&frozenxid_updated, &minmulti_updated, false);
+#else
 	TransactionId oldestXmin;
 	TransactionId freezeLimit;
 	MultiXactId multiXactCutoff;
 
 	/* initialize xids */
-#if PG_VERSION_NUM >= PG_VERSION_15
+#if (PG_VERSION_NUM >= PG_VERSION_15) && (PG_VERSION_NUM < PG_VERSION_16)
 	MultiXactId oldestMxact;
 	vacuum_set_xid_limits(rel,
 						  params->freeze_min_age,
@@ -1134,7 +1161,7 @@ columnar_vacuum_rel(Relation rel, VacuumParams *params,
 	 * just advance to the most aggressive value.
 	 */
 	TransactionId newRelFrozenXid = oldestXmin;
-#if PG_VERSION_NUM >= PG_VERSION_15
+#if (PG_VERSION_NUM >= PG_VERSION_15) && (PG_VERSION_NUM < PG_VERSION_16)
 	MultiXactId newRelminMxid = oldestMxact;
 #else
 	MultiXactId newRelminMxid = multiXactCutoff;
@@ -1145,7 +1172,7 @@ columnar_vacuum_rel(Relation rel, VacuumParams *params,
 	/* all visible pages are always 0 */
 	BlockNumber new_rel_allvisible = 0;
 
-#if PG_VERSION_NUM >= PG_VERSION_15
+#if (PG_VERSION_NUM >= PG_VERSION_15) && (PG_VERSION_NUM < PG_VERSION_16)
 	bool frozenxid_updated;
 	bool minmulti_updated;
 
@@ -1157,6 +1184,7 @@ columnar_vacuum_rel(Relation rel, VacuumParams *params,
 	vac_update_relstats(rel, new_rel_pages, new_live_tuples,
 						new_rel_allvisible, nindexes > 0,
 						newRelFrozenXid, newRelminMxid, false);
+#endif
 #endif
 
 	pgstat_report_vacuum(RelationGetRelid(rel),
@@ -1174,7 +1202,7 @@ static void
 LogRelationStats(Relation rel, int elevel)
 {
 	ListCell *stripeMetadataCell = NULL;
-	RelFileNode relfilenode = rel->rd_node;
+	RelFileLocator relfilelocator = RelationPhysicalIdentifier_compat(rel);
 	StringInfo infoBuf = makeStringInfo();
 
 	int compressionStats[COMPRESSION_COUNT] = { 0 };
@@ -1185,13 +1213,13 @@ LogRelationStats(Relation rel, int elevel)
 	uint64 droppedChunksWithData = 0;
 	uint64 totalDecompressedLength = 0;
 
-	List *stripeList = StripesForRelfilenode(relfilenode);
+	List *stripeList = StripesForRelfilelocator(relfilelocator);
 	int stripeCount = list_length(stripeList);
 
 	foreach(stripeMetadataCell, stripeList)
 	{
 		StripeMetadata *stripe = lfirst(stripeMetadataCell);
-		StripeSkipList *skiplist = ReadStripeSkipList(relfilenode, stripe->id,
+		StripeSkipList *skiplist = ReadStripeSkipList(relfilelocator, stripe->id,
 													  RelationGetDescr(rel),
 													  stripe->chunkCount,
 													  GetTransactionSnapshot());
@@ -1327,7 +1355,8 @@ TruncateColumnar(Relation rel, int elevel)
 	 * new stripes be added beyond highestPhysicalAddress while
 	 * we're truncating.
 	 */
-	uint64 newDataReservation = Max(GetHighestUsedAddress(rel->rd_node) + 1,
+	uint64 newDataReservation = Max(GetHighestUsedAddress(
+										RelationPhysicalIdentifier_compat(rel)) + 1,
 									ColumnarFirstLogicalOffset);
 
 	BlockNumber old_rel_pages = smgrnblocks(RelationGetSmgr(rel), MAIN_FORKNUM);
@@ -1395,15 +1424,32 @@ ConditionalLockRelationWithTimeout(Relation rel, LOCKMODE lockMode, int timeout,
 
 
 static bool
-columnar_scan_analyze_next_block(TableScanDesc scan, BlockNumber blockno,
+columnar_scan_analyze_next_block(TableScanDesc scan,
+#if PG_VERSION_NUM >= PG_VERSION_17
+								 ReadStream *stream)
+#else
+								 BlockNumber blockno,
 								 BufferAccessStrategy bstrategy)
+#endif
 {
 	/*
 	 * Our access method is not pages based, i.e. tuples are not confined
 	 * to pages boundaries. So not much to do here. We return true anyway
 	 * so acquire_sample_rows() in analyze.c would call our
 	 * columnar_scan_analyze_next_tuple() callback.
+	 * In PG17, we return false in case there is no buffer left, since
+	 * the outer loop changed in acquire_sample_rows(), and it is
+	 * expected for the scan_analyze_next_block function to check whether
+	 * there are any blocks left in the block sampler.
 	 */
+#if PG_VERSION_NUM >= PG_VERSION_17
+	Buffer buf = read_stream_next_buffer(stream, NULL);
+	if (!BufferIsValid(buf))
+	{
+		return false;
+	}
+	ReleaseBuffer(buf);
+#endif
 	return true;
 }
 
@@ -1476,8 +1522,7 @@ columnar_index_build_range_scan(Relation columnarRelation,
 	if (!IsBootstrapProcessingMode() && !indexInfo->ii_Concurrent)
 	{
 		/* ignore lazy VACUUM's */
-		OldestXmin = GetOldestNonRemovableTransactionId_compat(columnarRelation,
-															   PROCARRAY_FLAGS_VACUUM);
+		OldestXmin = GetOldestNonRemovableTransactionId(columnarRelation);
 	}
 
 	Snapshot snapshot = { 0 };
@@ -1805,8 +1850,8 @@ ColumnarReadMissingRowsIntoIndex(TableScanDesc scan, Relation indexRelation,
 		Relation columnarRelation = scan->rs_rd;
 		IndexUniqueCheck indexUniqueCheck =
 			indexInfo->ii_Unique ? UNIQUE_CHECK_YES : UNIQUE_CHECK_NO;
-		index_insert_compat(indexRelation, indexValues, indexNulls, columnarItemPointer,
-							columnarRelation, indexUniqueCheck, false, indexInfo);
+		index_insert(indexRelation, indexValues, indexNulls, columnarItemPointer,
+					 columnarRelation, indexUniqueCheck, false, indexInfo);
 
 		validateIndexState->tups_inserted += 1;
 	}
@@ -1835,8 +1880,8 @@ TupleSortSkipSmallerItemPointers(Tuplesortstate *tupleSort, ItemPointer targetIt
 		Datum *abbrev = NULL;
 		Datum tsDatum;
 		bool tsDatumIsNull;
-		if (!tuplesort_getdatum(tupleSort, forwardDirection, &tsDatum,
-								&tsDatumIsNull, abbrev))
+		if (!tuplesort_getdatum_compat(tupleSort, forwardDirection, false,
+									   &tsDatum, &tsDatumIsNull, abbrev))
 		{
 			ItemPointerSetInvalid(&tsItemPointerData);
 			break;
@@ -2010,7 +2055,7 @@ columnar_tableam_init()
 		&EnableVersionChecksColumnar,
 		true,
 		PGC_USERSET,
-		GUC_NO_SHOW_ALL,
+		GUC_NO_SHOW_ALL | GUC_NOT_IN_SAMPLE,
 		NULL, NULL, NULL);
 }
 
@@ -2077,12 +2122,13 @@ ColumnarTableDropHook(Oid relid)
 		 * tableam tables storage is managed by postgres.
 		 */
 		Relation rel = table_open(relid, AccessExclusiveLock);
-		RelFileNode relfilenode = rel->rd_node;
+		RelFileLocator relfilelocator = RelationPhysicalIdentifier_compat(rel);
 
-		DeleteMetadataRows(relfilenode);
+		DeleteMetadataRows(relfilelocator);
 		DeleteColumnarTableOptions(rel->rd_id, true);
 
-		MarkRelfilenodeDropped(relfilenode.relNode, GetCurrentSubTransactionId());
+		MarkRelfilenumberDropped(RelationPhysicalIdentifierNumber_compat(relfilelocator),
+								 GetCurrentSubTransactionId());
 
 		/* keep the lock since we did physical changes to the relation */
 		table_close(rel, NoLock);
@@ -2199,7 +2245,6 @@ ColumnarProcessAlterTable(AlterTableStmt *alterTableStmt, List **columnarOptions
 				columnarRangeVar = alterTableStmt->relation;
 			}
 		}
-#if PG_VERSION_NUM >= PG_VERSION_15
 		else if (alterTableCmd->subtype == AT_SetAccessMethod)
 		{
 			if (columnarRangeVar || *columnarOptions)
@@ -2210,14 +2255,15 @@ ColumnarProcessAlterTable(AlterTableStmt *alterTableStmt, List **columnarOptions
 									"Specify SET ACCESS METHOD before storage parameters, or use separate ALTER TABLE commands.")));
 			}
 
-			destIsColumnar = (strcmp(alterTableCmd->name, COLUMNAR_AM_NAME) == 0);
+			destIsColumnar = (strcmp(alterTableCmd->name ? alterTableCmd->name :
+									 default_table_access_method,
+									 COLUMNAR_AM_NAME) == 0);
 
 			if (srcIsColumnar && !destIsColumnar)
 			{
 				DeleteColumnarTableOptions(RelationGetRelid(rel), true);
 			}
 		}
-#endif /* PG_VERSION_15 */
 	}
 
 	relation_close(rel, NoLock);
@@ -2232,21 +2278,17 @@ ColumnarProcessAlterTable(AlterTableStmt *alterTableStmt, List **columnarOptions
 static void
 ColumnarProcessUtility(PlannedStmt *pstmt,
 					   const char *queryString,
-#if PG_VERSION_NUM >= PG_VERSION_14
 					   bool readOnlyTree,
-#endif
 					   ProcessUtilityContext context,
 					   ParamListInfo params,
 					   struct QueryEnvironment *queryEnv,
 					   DestReceiver *dest,
 					   QueryCompletion *completionTag)
 {
-#if PG_VERSION_NUM >= PG_VERSION_14
 	if (readOnlyTree)
 	{
 		pstmt = copyObject(pstmt);
 	}
-#endif
 
 	Node *parsetree = pstmt->utilityStmt;
 
@@ -2363,8 +2405,8 @@ ColumnarProcessUtility(PlannedStmt *pstmt,
 		CheckCitusColumnarAlterExtensionStmt(parsetree);
 	}
 
-	PrevProcessUtilityHook_compat(pstmt, queryString, false, context,
-								  params, queryEnv, dest, completionTag);
+	PrevProcessUtilityHook(pstmt, queryString, false, context,
+						   params, queryEnv, dest, completionTag);
 
 	if (columnarOptions != NIL)
 	{
@@ -2492,11 +2534,7 @@ static const TableAmRoutine columnar_am_methods = {
 	.tuple_get_latest_tid = columnar_get_latest_tid,
 	.tuple_tid_valid = columnar_tuple_tid_valid,
 	.tuple_satisfies_snapshot = columnar_tuple_satisfies_snapshot,
-#if PG_VERSION_NUM >= PG_VERSION_14
 	.index_delete_tuples = columnar_index_delete_tuples,
-#else
-	.compute_xid_horizon_for_tuples = columnar_compute_xid_horizon_for_tuples,
-#endif
 
 	.tuple_insert = columnar_tuple_insert,
 	.tuple_insert_speculative = columnar_tuple_insert_speculative,
@@ -2507,7 +2545,11 @@ static const TableAmRoutine columnar_am_methods = {
 	.tuple_lock = columnar_tuple_lock,
 	.finish_bulk_insert = columnar_finish_bulk_insert,
 
-	.relation_set_new_filenode = columnar_relation_set_new_filenode,
+#if PG_VERSION_NUM >= PG_VERSION_16
+	.relation_set_new_filelocator = columnar_relation_set_new_filelocator,
+#else
+	.relation_set_new_filenode = columnar_relation_set_new_filelocator,
+#endif
 	.relation_nontransactional_truncate = columnar_relation_nontransactional_truncate,
 	.relation_copy_data = columnar_relation_copy_data,
 	.relation_copy_for_cluster = columnar_relation_copy_for_cluster,
@@ -2568,8 +2610,13 @@ detoast_values(TupleDesc tupleDesc, Datum *orig_values, bool *isnull)
 			if (values == orig_values)
 			{
 				values = palloc(sizeof(Datum) * natts);
-				memcpy_s(values, sizeof(Datum) * natts,
-						 orig_values, sizeof(Datum) * natts);
+
+				/*
+				 * We use IGNORE-BANNED here since we don't want to limit
+				 * size of the buffer that holds the datum array to RSIZE_MAX
+				 * unnecessarily.
+				 */
+				memcpy(values, orig_values, sizeof(Datum) * natts); /* IGNORE-BANNED */
 			}
 
 			/* will be freed when per-tuple context is reset */
@@ -2600,21 +2647,12 @@ ColumnarCheckLogicalReplication(Relation rel)
 		return;
 	}
 
-#if PG_VERSION_NUM >= PG_VERSION_15
 	{
 		PublicationDesc pubdesc;
 
 		RelationBuildPublicationDesc(rel, &pubdesc);
 		pubActionInsert = pubdesc.pubactions.pubinsert;
 	}
-#else
-	if (rel->rd_pubactions == NULL)
-	{
-		GetRelationPublicationActions(rel);
-		Assert(rel->rd_pubactions != NULL);
-	}
-	pubActionInsert = rel->rd_pubactions->pubinsert;
-#endif
 
 	if (pubActionInsert)
 	{
@@ -2916,7 +2954,7 @@ MajorVersionsCompatibleColumnar(char *leftVersion, char *rightVersion)
 	}
 	else
 	{
-		rightComparisionLimit = strlen(leftVersion);
+		rightComparisionLimit = strlen(rightVersion);
 	}
 
 	/* we can error out early if hypens are not in the same position */
@@ -2991,6 +3029,8 @@ AvailableExtensionVersionColumnar(void)
 
 	ereport(ERROR, (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
 					errmsg("citus extension is not found")));
+
+	return NULL; /* keep compiler happy */
 }
 
 
@@ -3053,7 +3093,7 @@ DefElem *
 GetExtensionOption(List *extensionOptions, const char *defname)
 {
 	DefElem *defElement = NULL;
-	foreach_ptr(defElement, extensionOptions)
+	foreach_declared_ptr(defElement, extensionOptions)
 	{
 		if (IsA(defElement, DefElem) &&
 			strncmp(defElement->defname, defname, NAMEDATALEN) == 0)

src/backend/columnar/columnar_writer.c

@@ -16,28 +16,37 @@
 
 #include "postgres.h"
 
+#include "miscadmin.h"
 #include "safe_lib.h"
 
 #include "access/heapam.h"
 #include "access/nbtree.h"
 #include "catalog/pg_am.h"
-#include "miscadmin.h"
 #include "storage/fd.h"
 #include "storage/smgr.h"
 #include "utils/guc.h"
 #include "utils/memutils.h"
 #include "utils/rel.h"
-#include "utils/relfilenodemap.h"
+
+#include "pg_version_compat.h"
+#include "pg_version_constants.h"
 
 #include "columnar/columnar.h"
 #include "columnar/columnar_storage.h"
 #include "columnar/columnar_version_compat.h"
 
+#if PG_VERSION_NUM >= PG_VERSION_16
+#include "storage/relfilelocator.h"
+#include "utils/relfilenumbermap.h"
+#else
+#include "utils/relfilenodemap.h"
+#endif
+
 struct ColumnarWriteState
 {
 	TupleDesc tupleDescriptor;
 	FmgrInfo **comparisonFunctionArray;
-	RelFileNode relfilenode;
+	RelFileLocator relfilelocator;
 
 	MemoryContext stripeWriteContext;
 	MemoryContext perTupleContext;
@@ -84,7 +93,7 @@ static StringInfo CopyStringInfo(StringInfo sourceString);
  * data load operation.
  */
 ColumnarWriteState *
-ColumnarBeginWrite(RelFileNode relfilenode,
+ColumnarBeginWrite(RelFileLocator relfilelocator,
 				   ColumnarOptions options,
 				   TupleDesc tupleDescriptor)
 {
@@ -124,7 +133,7 @@ ColumnarBeginWrite(RelFileNode relfilenode,
 												options.chunkRowCount);
 
 	ColumnarWriteState *writeState = palloc0(sizeof(ColumnarWriteState));
-	writeState->relfilenode = relfilenode;
+	writeState->relfilelocator = relfilelocator;
 	writeState->options = options;
 	writeState->tupleDescriptor = CreateTupleDescCopy(tupleDescriptor);
 	writeState->comparisonFunctionArray = comparisonFunctionArray;
@@ -174,8 +183,10 @@ ColumnarWriteRow(ColumnarWriteState *writeState, Datum *columnValues, bool *colu
 		writeState->stripeSkipList = stripeSkipList;
 		writeState->compressionBuffer = makeStringInfo();
 
-		Oid relationId = RelidByRelfilenode(writeState->relfilenode.spcNode,
-											writeState->relfilenode.relNode);
+		Oid relationId = RelidByRelfilenumber(RelationTablespace_compat(
+												  writeState->relfilelocator),
+											  RelationPhysicalIdentifierNumber_compat(
+												  writeState->relfilelocator));
 		Relation relation = relation_open(relationId, NoLock);
 		writeState->emptyStripeReservation =
 			ReserveEmptyStripe(relation, columnCount, chunkRowCount,
@@ -393,8 +404,10 @@ FlushStripe(ColumnarWriteState *writeState)
 
 	elog(DEBUG1, "Flushing Stripe of size %d", stripeBuffers->rowCount);
 
-	Oid relationId = RelidByRelfilenode(writeState->relfilenode.spcNode,
-										writeState->relfilenode.relNode);
+	Oid relationId = RelidByRelfilenumber(RelationTablespace_compat(
+											  writeState->relfilelocator),
+										  RelationPhysicalIdentifierNumber_compat(
+											  writeState->relfilelocator));
 	Relation relation = relation_open(relationId, NoLock);
 
 	/*
@@ -486,10 +499,10 @@ FlushStripe(ColumnarWriteState *writeState)
 		}
 	}
 
-	SaveChunkGroups(writeState->relfilenode,
+	SaveChunkGroups(writeState->relfilelocator,
 					stripeMetadata->id,
 					writeState->chunkGroupRowCounts);
-	SaveStripeSkipList(writeState->relfilenode,
+	SaveStripeSkipList(writeState->relfilelocator,
 					   stripeMetadata->id,
 					   stripeSkipList, tupleDescriptor);
 
@@ -531,6 +544,9 @@ SerializeBoolArray(bool *boolArray, uint32 boolArrayLength)
 /*
  * SerializeSingleDatum serializes the given datum value and appends it to the
  * provided string info buffer.
+ *
+ * Since we don't want to limit datum buffer size to RSIZE_MAX unnecessarily,
+ * we use memcpy instead of memcpy_s several places in this function.
  */
 static void
 SerializeSingleDatum(StringInfo datumBuffer, Datum datum, bool datumTypeByValue,
@@ -552,15 +568,13 @@ SerializeSingleDatum(StringInfo datumBuffer, Datum datum, bool datumTypeByValue,
 		}
 		else
 		{
-			memcpy_s(currentDatumDataPointer, datumBuffer->maxlen - datumBuffer->len,
-					 DatumGetPointer(datum), datumTypeLength);
+			memcpy(currentDatumDataPointer, DatumGetPointer(datum), datumTypeLength); /* IGNORE-BANNED */
 		}
 	}
 	else
 	{
 		Assert(!datumTypeByValue);
-		memcpy_s(currentDatumDataPointer, datumBuffer->maxlen - datumBuffer->len,
-				 DatumGetPointer(datum), datumLength);
+		memcpy(currentDatumDataPointer, DatumGetPointer(datum), datumLength); /* IGNORE-BANNED */
 	}
 
 	datumBuffer->len += datumLengthAligned;
@@ -714,7 +728,12 @@ DatumCopy(Datum datum, bool datumTypeByValue, int datumTypeLength)
 	{
 		uint32 datumLength = att_addlength_datum(0, datumTypeLength, datum);
 		char *datumData = palloc0(datumLength);
-		memcpy_s(datumData, datumLength, DatumGetPointer(datum), datumLength);
+
+		/*
+		 * We use IGNORE-BANNED here since we don't want to limit datum size to
+		 * RSIZE_MAX unnecessarily.
+		 */
+		memcpy(datumData, DatumGetPointer(datum), datumLength); /* IGNORE-BANNED */
 
 		datumCopy = PointerGetDatum(datumData);
 	}
@@ -737,8 +756,12 @@ CopyStringInfo(StringInfo sourceString)
 		targetString->data = palloc0(sourceString->len);
 		targetString->len = sourceString->len;
 		targetString->maxlen = sourceString->len;
-		memcpy_s(targetString->data, sourceString->len,
-				 sourceString->data, sourceString->len);
+
+		/*
+		 * We use IGNORE-BANNED here since we don't want to limit string
+		 * buffer size to RSIZE_MAX unnecessarily.
+		 */
+		memcpy(targetString->data, sourceString->data, sourceString->len); /* IGNORE-BANNED */
 	}
 
 	return targetString;

src/backend/columnar/sql/citus_columnar--11.1-1--11.2-1.sql

@@ -0,0 +1,19 @@
+-- citus_columnar--11.1-1--11.2-1
+
+#include "udfs/columnar_ensure_am_depends_catalog/11.2-1.sql"
+
+DELETE FROM pg_depend
+WHERE classid = 'pg_am'::regclass::oid
+    AND objid IN (select oid from pg_am where amname = 'columnar')
+    AND objsubid = 0
+    AND refclassid = 'pg_class'::regclass::oid
+    AND refobjid IN (
+        'columnar_internal.stripe_first_row_number_idx'::regclass::oid,
+        'columnar_internal.chunk_group_pkey'::regclass::oid,
+        'columnar_internal.chunk_pkey'::regclass::oid,
+        'columnar_internal.options_pkey'::regclass::oid,
+        'columnar_internal.stripe_first_row_number_idx'::regclass::oid,
+        'columnar_internal.stripe_pkey'::regclass::oid
+    )
+    AND refobjsubid = 0
+    AND deptype = 'n';

src/backend/columnar/sql/citus_columnar--11.2-1--11.3-1.sql

@@ -0,0 +1 @@
+-- citus_columnar--11.2-1--11.3-1

src/backend/columnar/sql/citus_columnar--11.3-1--12.2-1.sql

@@ -0,0 +1 @@
+-- citus_columnar--11.3-1--12.2-1

src/backend/columnar/sql/columnar--10.2-2--10.2-3.sql

@@ -10,6 +10,17 @@
 --
 -- To do that, drop stripe_first_row_number_idx and create a unique
 -- constraint with the same name to keep the code change at minimum.
+--
+-- If we have a pg_depend entry for this index, we can not drop it as
+-- the extension depends on it. Remove the pg_depend entry if it exists.
+DELETE FROM pg_depend
+WHERE classid = 'pg_am'::regclass::oid
+    AND objid IN (select oid from pg_am where amname = 'columnar')
+    AND objsubid = 0
+    AND refclassid = 'pg_class'::regclass::oid
+    AND refobjid = 'columnar.stripe_first_row_number_idx'::regclass::oid
+    AND refobjsubid = 0
+    AND deptype = 'n';
 DROP INDEX columnar.stripe_first_row_number_idx;
 ALTER TABLE columnar.stripe ADD CONSTRAINT stripe_first_row_number_idx
 UNIQUE (storage_id, first_row_number);

src/backend/columnar/sql/downgrades/citus_columnar--11.2-1--11.1-1.sql

@@ -0,0 +1,4 @@
+-- citus_columnar--11.2-1--11.1-1
+
+-- Note that we intentionally do not re-insert the pg_depend records that we
+-- deleted via citus_columnar--11.1-1--11.2-1.sql.

src/backend/columnar/sql/downgrades/citus_columnar--11.3-1--11.2-1.sql

@@ -0,0 +1 @@
+-- citus_columnar--11.3-1--11.2-1

src/backend/columnar/sql/downgrades/citus_columnar--12.2-1--11.3-1.sql

@@ -0,0 +1 @@
+-- citus_columnar--12.2-1--11.3-1

src/backend/columnar/sql/downgrades/columnar--10.2-1--10.1-1.sql

@@ -8,5 +8,16 @@ DROP FUNCTION citus_internal.upgrade_columnar_storage(regclass);
 DROP FUNCTION citus_internal.downgrade_columnar_storage(regclass);
 
 -- drop "first_row_number" column and the index defined on it
+--
+-- If we have a pg_depend entry for this index, we can not drop it as
+-- the extension depends on it. Remove the pg_depend entry if it exists.
+DELETE FROM pg_depend
+WHERE classid = 'pg_am'::regclass::oid
+    AND objid IN (select oid from pg_am where amname = 'columnar')
+    AND objsubid = 0
+    AND refclassid = 'pg_class'::regclass::oid
+    AND refobjid = 'columnar.stripe_first_row_number_idx'::regclass::oid
+    AND refobjsubid = 0
+    AND deptype = 'n';
 DROP INDEX columnar.stripe_first_row_number_idx;
 ALTER TABLE columnar.stripe DROP COLUMN first_row_number;

src/backend/columnar/sql/downgrades/columnar--10.2-3--10.2-2.sql

@@ -1,4 +1,14 @@
 -- columnar--10.2-3--10.2-2.sql
-
+--
+-- If we have a pg_depend entry for this index, we can not drop it as
+-- the extension depends on it. Remove the pg_depend entry if it exists.
+DELETE FROM pg_depend
+WHERE classid = 'pg_am'::regclass::oid
+    AND objid IN (select oid from pg_am where amname = 'columnar')
+    AND objsubid = 0
+    AND refclassid = 'pg_class'::regclass::oid
+    AND refobjid = 'columnar.stripe_first_row_number_idx'::regclass::oid
+    AND refobjsubid = 0
+    AND deptype = 'n';
 ALTER TABLE columnar.stripe DROP CONSTRAINT stripe_first_row_number_idx;
 CREATE INDEX stripe_first_row_number_idx ON columnar.stripe USING BTREE(storage_id, first_row_number);

src/backend/columnar/sql/udfs/columnar_ensure_am_depends_catalog/11.2-1.sql

@@ -0,0 +1,43 @@
+CREATE OR REPLACE FUNCTION columnar_internal.columnar_ensure_am_depends_catalog()
+  RETURNS void
+  LANGUAGE plpgsql
+  SET search_path = pg_catalog
+AS $func$
+BEGIN
+  INSERT INTO pg_depend
+  WITH columnar_schema_members(relid) AS (
+    SELECT pg_class.oid AS relid FROM pg_class
+      WHERE relnamespace =
+            COALESCE(
+	       (SELECT pg_namespace.oid FROM pg_namespace WHERE nspname = 'columnar_internal'),
+	       (SELECT pg_namespace.oid FROM pg_namespace WHERE nspname = 'columnar')
+	    )
+        AND relname IN ('chunk',
+                        'chunk_group',
+                        'options',
+                        'storageid_seq',
+                        'stripe')
+  )
+  SELECT -- Define a dependency edge from "columnar table access method" ..
+         'pg_am'::regclass::oid as classid,
+         (select oid from pg_am where amname = 'columnar') as objid,
+         0 as objsubid,
+         -- ... to some objects registered as regclass and that lives in
+         -- "columnar" schema. That contains catalog tables and the sequences
+         -- created in "columnar" schema.
+         --
+         -- Given the possibility of user might have created their own objects
+         -- in columnar schema, we explicitly specify list of objects that we
+         -- are interested in.
+         'pg_class'::regclass::oid as refclassid,
+         columnar_schema_members.relid as refobjid,
+         0 as refobjsubid,
+         'n' as deptype
+  FROM columnar_schema_members
+  -- Avoid inserting duplicate entries into pg_depend.
+  EXCEPT TABLE pg_depend;
+END;
+$func$;
+COMMENT ON FUNCTION columnar_internal.columnar_ensure_am_depends_catalog()
+  IS 'internal function responsible for creating dependencies from columnar '
+     'table access method to the rel objects in columnar schema';

src/backend/columnar/sql/udfs/columnar_ensure_am_depends_catalog/latest.sql

@@ -1,4 +1,4 @@
-CREATE OR REPLACE FUNCTION citus_internal.columnar_ensure_am_depends_catalog()
+CREATE OR REPLACE FUNCTION columnar_internal.columnar_ensure_am_depends_catalog()
   RETURNS void
   LANGUAGE plpgsql
   SET search_path = pg_catalog
@@ -14,22 +14,17 @@ BEGIN
 	    )
         AND relname IN ('chunk',
                         'chunk_group',
-                        'chunk_group_pkey',
-                        'chunk_pkey',
                         'options',
-                        'options_pkey',
                         'storageid_seq',
-                        'stripe',
-                        'stripe_first_row_number_idx',
-                        'stripe_pkey')
+                        'stripe')
   )
   SELECT -- Define a dependency edge from "columnar table access method" ..
          'pg_am'::regclass::oid as classid,
          (select oid from pg_am where amname = 'columnar') as objid,
          0 as objsubid,
-         -- ... to each object that is registered to pg_class and that lives
-         -- in "columnar" schema. That contains catalog tables, indexes
-         -- created on them and the sequences created in "columnar" schema.
+         -- ... to some objects registered as regclass and that lives in
+         -- "columnar" schema. That contains catalog tables and the sequences
+         -- created in "columnar" schema.
          --
          -- Given the possibility of user might have created their own objects
          -- in columnar schema, we explicitly specify list of objects that we
@@ -43,6 +38,6 @@ BEGIN
   EXCEPT TABLE pg_depend;
 END;
 $func$;
-COMMENT ON FUNCTION citus_internal.columnar_ensure_am_depends_catalog()
+COMMENT ON FUNCTION columnar_internal.columnar_ensure_am_depends_catalog()
   IS 'internal function responsible for creating dependencies from columnar '
      'table access method to the rel objects in columnar schema';

src/backend/columnar/write_state_management.c

@@ -1,21 +1,17 @@
 
-#include "citus_version.h"
-
-#include "postgres.h"
-#include "columnar/columnar.h"
-
-
 #include <math.h>
 
+#include "postgres.h"
+
 #include "miscadmin.h"
+#include "pgstat.h"
 
 #include "access/genam.h"
 #include "access/heapam.h"
+#include "access/heaptoast.h"
 #include "access/multixact.h"
 #include "access/rewriteheap.h"
 #include "access/tsmapi.h"
-#include "access/heaptoast.h"
-#include "common/hashfn.h"
 #include "access/xact.h"
 #include "catalog/catalog.h"
 #include "catalog/index.h"
@@ -26,13 +22,12 @@
 #include "catalog/storage_xlog.h"
 #include "commands/progress.h"
 #include "commands/vacuum.h"
+#include "common/hashfn.h"
 #include "executor/executor.h"
 #include "nodes/makefuncs.h"
 #include "optimizer/plancat.h"
-#include "pgstat.h"
 #include "storage/bufmgr.h"
 #include "storage/bufpage.h"
-#include "storage/bufmgr.h"
 #include "storage/lmgr.h"
 #include "storage/predicate.h"
 #include "storage/procarray.h"
@@ -43,6 +38,10 @@
 #include "utils/rel.h"
 #include "utils/syscache.h"
 
+#include "citus_version.h"
+#include "pg_version_compat.h"
+
+#include "columnar/columnar.h"
 #include "columnar/columnar_customscan.h"
 #include "columnar/columnar_tableam.h"
 #include "columnar/columnar_version_compat.h"
@@ -77,7 +76,7 @@ typedef struct SubXidWriteState
 typedef struct WriteStateMapEntry
 {
 	/* key of the entry */
-	Oid relfilenode;
+	RelFileNumber relfilenumber;
 
 	/*
 	 * If a table is dropped, we set dropped to true and set dropSubXid to the
@@ -113,6 +112,7 @@ CleanupWriteStateMap(void *arg)
 
 ColumnarWriteState *
 columnar_init_write_state(Relation relation, TupleDesc tupdesc,
+						  Oid tupSlotRelationId,
 						  SubTransactionId currentSubXid)
 {
 	bool found;
@@ -131,7 +131,7 @@ columnar_init_write_state(Relation relation, TupleDesc tupdesc,
 		HASHCTL info;
 		uint32 hashFlags = (HASH_ELEM | HASH_FUNCTION | HASH_CONTEXT);
 		memset(&info, 0, sizeof(info));
-		info.keysize = sizeof(Oid);
+		info.keysize = sizeof(RelFileNumber);
 		info.hash = oid_hash;
 		info.entrysize = sizeof(WriteStateMapEntry);
 		info.hcxt = WriteStateContext;
@@ -145,7 +145,10 @@ columnar_init_write_state(Relation relation, TupleDesc tupdesc,
 		MemoryContextRegisterResetCallback(WriteStateContext, &cleanupCallback);
 	}
 
-	WriteStateMapEntry *hashEntry = hash_search(WriteStateMap, &relation->rd_node.relNode,
+	WriteStateMapEntry *hashEntry = hash_search(WriteStateMap,
+												&RelationPhysicalIdentifierNumber_compat(
+													RelationPhysicalIdentifier_compat(
+														relation)),
 												HASH_ENTER, &found);
 	if (!found)
 	{
@@ -176,10 +179,20 @@ columnar_init_write_state(Relation relation, TupleDesc tupdesc,
 	MemoryContext oldContext = MemoryContextSwitchTo(WriteStateContext);
 
 	ColumnarOptions columnarOptions = { 0 };
-	ReadColumnarOptions(relation->rd_id, &columnarOptions);
+
+	/*
+	 * In case of a table rewrite, we need to fetch table options based on the
+	 * relation id of the source tuple slot.
+	 *
+	 * For this reason, we always pass tupSlotRelationId here; which should be
+	 * same as the target table if the write operation is not related to a table
+	 * rewrite etc.
+	 */
+	ReadColumnarOptions(tupSlotRelationId, &columnarOptions);
 
 	SubXidWriteState *stackEntry = palloc0(sizeof(SubXidWriteState));
-	stackEntry->writeState = ColumnarBeginWrite(relation->rd_node,
+	stackEntry->writeState = ColumnarBeginWrite(RelationPhysicalIdentifier_compat(
+													relation),
 												columnarOptions,
 												tupdesc);
 	stackEntry->subXid = currentSubXid;
@@ -196,14 +209,16 @@ columnar_init_write_state(Relation relation, TupleDesc tupdesc,
  * Flushes pending writes for given relfilenode in the given subtransaction.
  */
 void
-FlushWriteStateForRelfilenode(Oid relfilenode, SubTransactionId currentSubXid)
+FlushWriteStateForRelfilenumber(RelFileNumber relfilenumber,
+								SubTransactionId currentSubXid)
 {
 	if (WriteStateMap == NULL)
 	{
 		return;
 	}
 
-	WriteStateMapEntry *entry = hash_search(WriteStateMap, &relfilenode, HASH_FIND, NULL);
+	WriteStateMapEntry *entry = hash_search(WriteStateMap, &relfilenumber, HASH_FIND,
+											NULL);
 
 	Assert(!entry || !entry->dropped);
 
@@ -310,14 +325,14 @@ DiscardWriteStateForAllRels(SubTransactionId currentSubXid, SubTransactionId par
  * Called when the given relfilenode is dropped.
  */
 void
-MarkRelfilenodeDropped(Oid relfilenode, SubTransactionId currentSubXid)
+MarkRelfilenumberDropped(RelFileNumber relfilenumber, SubTransactionId currentSubXid)
 {
 	if (WriteStateMap == NULL)
 	{
 		return;
 	}
 
-	WriteStateMapEntry *entry = hash_search(WriteStateMap, &relfilenode, HASH_FIND,
+	WriteStateMapEntry *entry = hash_search(WriteStateMap, &relfilenumber, HASH_FIND,
 											NULL);
 	if (!entry || entry->dropped)
 	{
@@ -333,11 +348,11 @@ MarkRelfilenodeDropped(Oid relfilenode, SubTransactionId currentSubXid)
  * Called when the given relfilenode is dropped in non-transactional TRUNCATE.
  */
 void
-NonTransactionDropWriteState(Oid relfilenode)
+NonTransactionDropWriteState(RelFileNumber relfilenumber)
 {
 	if (WriteStateMap)
 	{
-		hash_search(WriteStateMap, &relfilenode, HASH_REMOVE, false);
+		hash_search(WriteStateMap, &relfilenumber, HASH_REMOVE, false);
 	}
 }
 
@@ -346,14 +361,16 @@ NonTransactionDropWriteState(Oid relfilenode)
  * Returns true if there are any pending writes in upper transactions.
  */
 bool
-PendingWritesInUpperTransactions(Oid relfilenode, SubTransactionId currentSubXid)
+PendingWritesInUpperTransactions(RelFileNumber relfilenumber,
+								 SubTransactionId currentSubXid)
 {
 	if (WriteStateMap == NULL)
 	{
 		return false;
 	}
 
-	WriteStateMapEntry *entry = hash_search(WriteStateMap, &relfilenode, HASH_FIND, NULL);
+	WriteStateMapEntry *entry = hash_search(WriteStateMap, &relfilenumber, HASH_FIND,
+											NULL);
 
 	if (entry && entry->writeStateStack != NULL)
 	{

src/backend/distributed/Makefile

@@ -18,7 +18,7 @@ generated_downgrade_sql_files += $(patsubst %,$(citus_abs_srcdir)/build/sql/%,$(
 DATA_built = $(generated_sql_files)
 
 # directories with source files
-SUBDIRS = . commands connection ddl deparser executor metadata operations planner progress relay safeclib shardsplit test transaction utils worker
+SUBDIRS = . commands connection ddl deparser executor metadata operations planner progress relay safeclib shardsplit stats test transaction utils worker clock
 # enterprise modules
 SUBDIRS += replication
 
@@ -32,7 +32,12 @@ OBJS += \
 	$(patsubst $(citus_abs_srcdir)/%.c,%.o,$(foreach dir,$(SUBDIRS), $(sort $(wildcard $(citus_abs_srcdir)/$(dir)/*.c))))
 
 # be explicit about the default target
-all:
+.PHONY: cdc
+
+all: cdc
+
+cdc:
+	$(MAKE) -C cdc all
 
 NO_PGXS = 1
 
@@ -79,13 +84,22 @@ ifneq (,$(SQL_Po_files))
 include $(SQL_Po_files)
 endif
 
-.PHONY: clean-full install install-downgrades install-all
+
+.PHONY: clean-full install install-downgrades install-all install-cdc clean-cdc
+
+clean: clean-cdc
+
+clean-cdc:
+	$(MAKE) -C cdc clean
 
 cleanup-before-install:
 	rm -f $(DESTDIR)$(datadir)/$(datamoduledir)/citus.control
 	rm -f $(DESTDIR)$(datadir)/$(datamoduledir)/citus--*
 
-install: cleanup-before-install
+install: cleanup-before-install install-cdc
+
+install-cdc:
+	$(MAKE) -C cdc install
 
 # install and install-downgrades should be run sequentially
 install-all: install
@@ -96,4 +110,5 @@ install-downgrades: $(generated_downgrade_sql_files)
 
 clean-full:
 	$(MAKE) clean
+	$(MAKE) -C cdc clean-full
 	rm -rf $(safestringlib_builddir)

src/backend/distributed/cdc/Makefile

@@ -0,0 +1,45 @@
+citus_top_builddir = ../../../..
+include $(citus_top_builddir)/Makefile.global
+
+citus_subdir = src/backend/distributed/cdc
+SRC_DIR = $(citus_abs_top_srcdir)/$(citus_subdir)
+
+#List of supported based decoders. Add new decoders here.
+cdc_base_decoders :=pgoutput wal2json
+
+all: build-cdc-decoders
+
+copy-decoder-files-to-build-dir:
+	$(eval DECODER_BUILD_DIR=build-cdc-$(DECODER))
+	mkdir -p $(DECODER_BUILD_DIR)
+	@for file in $(SRC_DIR)/*.c $(SRC_DIR)/*.h; do \
+		if [ -f $$file ]; then \
+			if [ -f $(DECODER_BUILD_DIR)/$$(basename $$file) ]; then \
+				if ! diff -q $$file $(DECODER_BUILD_DIR)/$$(basename $$file); then \
+					cp $$file $(DECODER_BUILD_DIR)/$$(basename $$file); \
+				fi \
+			else \
+				cp $$file $(DECODER_BUILD_DIR)/$$(basename $$file); \
+			fi \
+		fi \
+	done
+	cp $(SRC_DIR)/Makefile.decoder $(DECODER_BUILD_DIR)/Makefile
+
+build-cdc-decoders:
+	$(foreach base_decoder,$(cdc_base_decoders),$(MAKE) DECODER=$(base_decoder) build-cdc-decoder;)
+
+install-cdc-decoders:
+	$(foreach base_decoder,$(cdc_base_decoders),$(MAKE) DECODER=$(base_decoder) -C build-cdc-$(base_decoder) install;)
+
+clean-cdc-decoders:
+	$(foreach base_decoder,$(cdc_base_decoders),rm -rf build-cdc-$(base_decoder);)
+
+
+build-cdc-decoder:
+	$(MAKE) DECODER=$(DECODER) copy-decoder-files-to-build-dir
+	$(MAKE) DECODER=$(DECODER) -C build-cdc-$(DECODER)
+
+install: install-cdc-decoders
+
+clean: clean-cdc-decoders
+

src/backend/distributed/cdc/Makefile.decoder

@@ -0,0 +1,19 @@
+MODULE_big = citus_$(DECODER)
+
+citus_decoders_dir = $(DESTDIR)$(pkglibdir)/citus_decoders
+
+citus_top_builddir = ../../../../..
+citus_subdir = src/backend/distributed/cdc/cdc_$(DECODER)
+
+OBJS += cdc_decoder.o cdc_decoder_utils.o
+
+include $(citus_top_builddir)/Makefile.global
+
+override CFLAGS += -DDECODER=\"$(DECODER)\" -I$(citus_abs_top_srcdir)/include
+override CPPFLAGS += -DDECODER=\"$(DECODER)\" -I$(citus_abs_top_srcdir)/include
+
+install: install-cdc
+
+install-cdc:
+	mkdir -p '$(citus_decoders_dir)'
+	$(INSTALL_SHLIB) citus_$(DECODER)$(DLSUFFIX) '$(citus_decoders_dir)/$(DECODER)$(DLSUFFIX)'

src/backend/distributed/cdc/cdc_decoder.c

@@ -0,0 +1,573 @@
+/*-------------------------------------------------------------------------
+ *
+ * cdc_decoder.c
+ *		CDC Decoder plugin for Citus
+ *
+ * Copyright (c) Citus Data, Inc.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#include "postgres.h"
+
+#include "cdc_decoder_utils.h"
+#include "fmgr.h"
+
+#include "access/genam.h"
+#include "catalog/pg_namespace.h"
+#include "catalog/pg_publication.h"
+#include "commands/extension.h"
+#include "common/hashfn.h"
+#include "utils/lsyscache.h"
+#include "utils/rel.h"
+#include "utils/typcache.h"
+
+#include "pg_version_constants.h"
+
+PG_MODULE_MAGIC;
+
+extern void _PG_output_plugin_init(OutputPluginCallbacks *cb);
+static LogicalDecodeChangeCB ouputPluginChangeCB;
+
+static void InitShardToDistributedTableMap(void);
+
+static void PublishDistributedTableChanges(LogicalDecodingContext *ctx,
+										   ReorderBufferTXN *txn,
+										   Relation relation,
+										   ReorderBufferChange *change);
+
+
+static bool replication_origin_filter_cb(LogicalDecodingContext *ctx, RepOriginId
+										 origin_id);
+
+static void TranslateChangesIfSchemaChanged(Relation relation, Relation targetRelation,
+											ReorderBufferChange *change);
+
+static void TranslateAndPublishRelationForCDC(LogicalDecodingContext *ctx,
+											  ReorderBufferTXN *txn,
+											  Relation relation,
+											  ReorderBufferChange *change, Oid shardId,
+											  Oid targetRelationid);
+
+typedef struct
+{
+	uint64 shardId;
+	Oid distributedTableId;
+	bool isReferenceTable;
+	bool isNull;
+} ShardIdHashEntry;
+
+static HTAB *shardToDistributedTableMap = NULL;
+
+static void cdc_change_cb(LogicalDecodingContext *ctx, ReorderBufferTXN *txn,
+						  Relation relation, ReorderBufferChange *change);
+
+
+/* build time macro for base decoder plugin name for CDC and Shard Split. */
+#ifndef DECODER
+#define DECODER "pgoutput"
+#endif
+
+#define DECODER_INIT_FUNCTION_NAME "_PG_output_plugin_init"
+
+#define CITUS_SHARD_TRANSFER_SLOT_PREFIX "citus_shard_"
+#define CITUS_SHARD_TRANSFER_SLOT_PREFIX_SIZE (sizeof(CITUS_SHARD_TRANSFER_SLOT_PREFIX) - \
+											   1)
+
+/*
+ * Postgres uses 'pgoutput' as default plugin for logical replication.
+ * We want to reuse Postgres pgoutput's functionality as much as possible.
+ * Hence we load all the functions of this plugin and override as required.
+ */
+void
+_PG_output_plugin_init(OutputPluginCallbacks *cb)
+{
+	elog(LOG, "Initializing CDC decoder");
+
+	/*
+	 * We build custom .so files whose name matches common decoders (pgoutput, wal2json)
+	 * and place them in $libdir/citus_decoders/ such that administrators can configure
+	 * dynamic_library_path to include this directory, and users can then use the
+	 * regular decoder names when creating replications slots.
+	 *
+	 * To load the original decoder, we need to remove citus_decoders/ from the
+	 * dynamic_library_path.
+	 */
+	char *originalDLP = Dynamic_library_path;
+	Dynamic_library_path = RemoveCitusDecodersFromPaths(Dynamic_library_path);
+
+	LogicalOutputPluginInit plugin_init =
+		(LogicalOutputPluginInit) (void *)
+		load_external_function(DECODER,
+							   DECODER_INIT_FUNCTION_NAME,
+							   false, NULL);
+
+	if (plugin_init == NULL)
+	{
+		elog(ERROR, "output plugins have to declare the _PG_output_plugin_init symbol");
+	}
+
+	/* in case this session is used for different replication slots */
+	Dynamic_library_path = originalDLP;
+
+	/* ask the output plugin to fill the callback struct */
+	plugin_init(cb);
+
+	/* Initialize the Shard Id to Distributed Table id mapping hash table.*/
+	InitShardToDistributedTableMap();
+
+	/* actual pgoutput callback function will be called  */
+	ouputPluginChangeCB = cb->change_cb;
+	cb->change_cb = cdc_change_cb;
+	cb->filter_by_origin_cb = replication_origin_filter_cb;
+}
+
+
+/*
+ *  Check if the replication slot is for Shard transfer by checking for prefix.
+ */
+inline static
+bool
+IsShardTransferSlot(char *replicationSlotName)
+{
+	return strncmp(replicationSlotName, CITUS_SHARD_TRANSFER_SLOT_PREFIX,
+				   CITUS_SHARD_TRANSFER_SLOT_PREFIX_SIZE) == 0;
+}
+
+
+/*
+ * shard_split_and_cdc_change_cb function emits the incoming tuple change
+ * to the appropriate destination shard.
+ */
+static void
+cdc_change_cb(LogicalDecodingContext *ctx, ReorderBufferTXN *txn,
+			  Relation relation, ReorderBufferChange *change)
+{
+	/*
+	 * If Citus has not been loaded yet, pass the changes
+	 * through to the undrelying decoder plugin.
+	 */
+	if (!CdcCitusHasBeenLoaded())
+	{
+		ouputPluginChangeCB(ctx, txn, relation, change);
+		return;
+	}
+
+	/* check if the relation is publishable.*/
+	if (!is_publishable_relation(relation))
+	{
+		return;
+	}
+
+	char *replicationSlotName = ctx->slot->data.name.data;
+	if (replicationSlotName == NULL)
+	{
+		elog(ERROR, "Replication slot name is NULL!");
+		return;
+	}
+
+	/* If the slot is for internal shard operations, call the base plugin's call back. */
+	if (IsShardTransferSlot(replicationSlotName))
+	{
+		ouputPluginChangeCB(ctx, txn, relation, change);
+		return;
+	}
+
+	/* Transalate the changes from shard to distributes table and publish. */
+	PublishDistributedTableChanges(ctx, txn, relation, change);
+}
+
+
+/*
+ * InitShardToDistributedTableMap initializes the hash table that is used to
+ * translate the changes in the shard table to the changes in the distributed table.
+ */
+static void
+InitShardToDistributedTableMap()
+{
+	HASHCTL info;
+	memset(&info, 0, sizeof(info));
+	info.keysize = sizeof(uint64);
+	info.entrysize = sizeof(ShardIdHashEntry);
+	info.hash = tag_hash;
+	info.hcxt = CurrentMemoryContext;
+
+	int hashFlags = (HASH_ELEM | HASH_CONTEXT | HASH_FUNCTION);
+	shardToDistributedTableMap = hash_create("CDC Decoder translation hash table", 1024,
+											 &info, hashFlags);
+}
+
+
+/*
+ * AddShardIdToHashTable adds the shardId to the hash table.
+ */
+static Oid
+AddShardIdToHashTable(uint64 shardId, ShardIdHashEntry *entry)
+{
+	entry->shardId = shardId;
+	entry->distributedTableId = CdcLookupShardRelationFromCatalog(shardId, true);
+	entry->isReferenceTable = CdcIsReferenceTableViaCatalog(entry->distributedTableId);
+	return entry->distributedTableId;
+}
+
+
+static Oid
+LookupDistributedTableIdForShardId(uint64 shardId, bool *isReferenceTable)
+{
+	bool found;
+	Oid distributedTableId = InvalidOid;
+	ShardIdHashEntry *entry = (ShardIdHashEntry *) hash_search(shardToDistributedTableMap,
+															   &shardId,
+															   HASH_ENTER,
+															   &found);
+	if (found)
+	{
+		distributedTableId = entry->distributedTableId;
+	}
+	else
+	{
+		distributedTableId = AddShardIdToHashTable(shardId, entry);
+	}
+	*isReferenceTable = entry->isReferenceTable;
+	return distributedTableId;
+}
+
+
+/*
+ * replication_origin_filter_cb call back function filters out publication of changes
+ * originated from any other node other than the current node. This is
+ * identified by the "origin_id" of the changes. The origin_id is set to
+ * a non-zero value in the origin node as part of WAL replication for internal
+ * operations like shard split/moves/create_distributed_table etc.
+ */
+static bool
+replication_origin_filter_cb(LogicalDecodingContext *ctx, RepOriginId origin_id)
+{
+	return  (origin_id != InvalidRepOriginId);
+}
+
+
+/*
+ * This function is responsible for translating the changes in the shard table to
+ * the changes in the shell table and publishing the changes as a change to the
+ * distributed table so that CDD clients are not aware of the shard tables. It also
+ * handles schema changes to the distributed table.
+ */
+static void
+TranslateAndPublishRelationForCDC(LogicalDecodingContext *ctx, ReorderBufferTXN *txn,
+								  Relation relation, ReorderBufferChange *change, Oid
+								  shardId, Oid targetRelationid)
+{
+	/* Get the distributed table's relation for this shard.*/
+	Relation targetRelation = RelationIdGetRelation(targetRelationid);
+
+	/*
+	 * Check if there has been a schema change (such as a dropped column), by comparing
+	 * the number of attributes in the shard table and the shell table.
+	 */
+	TranslateChangesIfSchemaChanged(relation, targetRelation, change);
+
+	/*
+	 * Publish the change to the shard table as the change in the distributed table,
+	 * so that the CDC client can see the change in the distributed table,
+	 * instead of the shard table, by calling the pgoutput's callback function.
+	 */
+	ouputPluginChangeCB(ctx, txn, targetRelation, change);
+	RelationClose(targetRelation);
+}
+
+
+/*
+ * PublishChangesIfCdcSlot checks if the current slot is a CDC slot. If so, it publishes
+ * the changes as the change for the distributed table instead of shard.
+ * If not, it returns false. It also skips the Citus metadata tables.
+ */
+static void
+PublishDistributedTableChanges(LogicalDecodingContext *ctx, ReorderBufferTXN *txn,
+							   Relation relation, ReorderBufferChange *change)
+{
+	char *shardRelationName = RelationGetRelationName(relation);
+
+	/* Skip publishing CDC changes for any system relations in pg_catalog*/
+	if (relation->rd_rel->relnamespace == PG_CATALOG_NAMESPACE)
+	{
+		return;
+	}
+
+	/* Check if the relation is a distributed table by checking for shard name.	*/
+	uint64 shardId = CdcExtractShardIdFromTableName(shardRelationName, true);
+
+	/* If this relation is not distributed, call the pgoutput's callback and return. */
+	if (shardId == INVALID_SHARD_ID)
+	{
+		ouputPluginChangeCB(ctx, txn, relation, change);
+		return;
+	}
+
+	bool isReferenceTable = false;
+	Oid distRelationId = LookupDistributedTableIdForShardId(shardId, &isReferenceTable);
+	if (distRelationId == InvalidOid)
+	{
+		ouputPluginChangeCB(ctx, txn, relation, change);
+		return;
+	}
+
+	/* Publish changes for reference table only from the coordinator node. */
+	if (isReferenceTable && !CdcIsCoordinator())
+	{
+		return;
+	}
+
+	/* translate and publish from shard relation to distributed table relation for CDC. */
+	TranslateAndPublishRelationForCDC(ctx, txn, relation, change, shardId,
+									  distRelationId);
+}
+
+
+/*
+ * GetTupleForTargetSchemaForCdc returns a heap tuple with the data from sourceRelationTuple
+ * to match the schema in targetRelDesc. Either or both source and target relations may have
+ * dropped columns. This function handles it by adding NULL values for dropped columns in
+ * target relation and skipping dropped columns in source relation. It returns a heap tuple
+ * adjusted to the current schema of the target relation.
+ */
+static HeapTuple
+GetTupleForTargetSchemaForCdc(HeapTuple sourceRelationTuple,
+							  TupleDesc sourceRelDesc,
+							  TupleDesc targetRelDesc)
+{
+	/* Allocate memory for sourceValues and sourceNulls arrays. */
+	Datum *sourceValues = (Datum *) palloc0(sourceRelDesc->natts * sizeof(Datum));
+	bool *sourceNulls = (bool *) palloc0(sourceRelDesc->natts * sizeof(bool));
+
+	/* Deform the source tuple to sourceValues and sourceNulls arrays. */
+	heap_deform_tuple(sourceRelationTuple, sourceRelDesc, sourceValues,
+					  sourceNulls);
+
+	/* This is the next field to Read in the source relation */
+	uint32 sourceIndex = 0;
+	uint32 targetIndex = 0;
+
+	/* Allocate memory for sourceValues and sourceNulls arrays. */
+	Datum *targetValues = (Datum *) palloc0(targetRelDesc->natts * sizeof(Datum));
+	bool *targetNulls = (bool *) palloc0(targetRelDesc->natts * sizeof(bool));
+
+	/* Loop through all source and target attributes one by one and handle any dropped attributes.*/
+	while (targetIndex < targetRelDesc->natts)
+	{
+		/* If this target attribute has been dropped, add a NULL attribute in targetValues and continue.*/
+		if (TupleDescAttr(targetRelDesc, targetIndex)->attisdropped)
+		{
+			Datum nullDatum = (Datum) 0;
+			targetValues[targetIndex] = nullDatum;
+			targetNulls[targetIndex] = true;
+			targetIndex++;
+		}
+
+		/* If this source attribute has been dropped, just skip this source attribute.*/
+		else if (TupleDescAttr(sourceRelDesc, sourceIndex)->attisdropped)
+		{
+			sourceIndex++;
+			continue;
+		}
+
+		/* If both source and target attributes are not dropped, add the attribute field to targetValues. */
+		else if (sourceIndex < sourceRelDesc->natts)
+		{
+			targetValues[targetIndex] = sourceValues[sourceIndex];
+			targetNulls[targetIndex] = sourceNulls[sourceIndex];
+			sourceIndex++;
+			targetIndex++;
+		}
+		else
+		{
+			/* If there are no more source fields, add a NULL field in targetValues. */
+			Datum nullDatum = (Datum) 0;
+			targetValues[targetIndex] = nullDatum;
+			targetNulls[targetIndex] = true;
+			targetIndex++;
+		}
+	}
+
+	/* Form a new tuple from the target values created by the above loop. */
+	HeapTuple targetRelationTuple = heap_form_tuple(targetRelDesc, targetValues,
+													targetNulls);
+	return targetRelationTuple;
+}
+
+
+/* HasSchemaChanged function returns if there any schema changes between source and target relations.*/
+static bool
+HasSchemaChanged(TupleDesc sourceRelationDesc, TupleDesc targetRelationDesc)
+{
+	bool hasSchemaChanged = (sourceRelationDesc->natts != targetRelationDesc->natts);
+	if (hasSchemaChanged)
+	{
+		return true;
+	}
+
+	for (uint32 i = 0; i < sourceRelationDesc->natts; i++)
+	{
+		if (TupleDescAttr(sourceRelationDesc, i)->attisdropped ||
+			TupleDescAttr(targetRelationDesc, i)->attisdropped)
+		{
+			hasSchemaChanged = true;
+			break;
+		}
+	}
+
+	return hasSchemaChanged;
+}
+
+
+/*
+ * TranslateChangesIfSchemaChanged translates the tuples ReorderBufferChange
+ * if there is a schema change between source and target relations.
+ */
+static void
+TranslateChangesIfSchemaChanged(Relation sourceRelation, Relation targetRelation,
+								ReorderBufferChange *change)
+{
+	TupleDesc sourceRelationDesc = RelationGetDescr(sourceRelation);
+	TupleDesc targetRelationDesc = RelationGetDescr(targetRelation);
+
+	/* if there are no changes between source and target relations, return. */
+	if (!HasSchemaChanged(sourceRelationDesc, targetRelationDesc))
+	{
+		return;
+	}
+
+#if PG_VERSION_NUM >= PG_VERSION_17
+
+	/* Check the ReorderBufferChange's action type and handle them accordingly.*/
+	switch (change->action)
+	{
+		case REORDER_BUFFER_CHANGE_INSERT:
+		{
+			/* For insert action, only new tuple should always be translated*/
+			HeapTuple sourceRelationNewTuple = change->data.tp.newtuple;
+			HeapTuple targetRelationNewTuple = GetTupleForTargetSchemaForCdc(
+				sourceRelationNewTuple, sourceRelationDesc, targetRelationDesc);
+			change->data.tp.newtuple = targetRelationNewTuple;
+			break;
+		}
+
+		/*
+		 * For update changes both old and new tuples need to be translated for target relation
+		 * if the REPLICA IDENTITY is set to FULL. Otherwise, only the new tuple needs to be
+		 * translated for target relation.
+		 */
+		case REORDER_BUFFER_CHANGE_UPDATE:
+		{
+			/* For update action, new tuple should always be translated*/
+			/* Get the new tuple from the ReorderBufferChange, and translate it to target relation. */
+			HeapTuple sourceRelationNewTuple = change->data.tp.newtuple;
+			HeapTuple targetRelationNewTuple = GetTupleForTargetSchemaForCdc(
+				sourceRelationNewTuple, sourceRelationDesc, targetRelationDesc);
+			change->data.tp.newtuple = targetRelationNewTuple;
+
+			/*
+			 * Format oldtuple according to the target relation. If the column values of replica
+			 * identiy change, then the old tuple is non-null and needs to be formatted according
+			 * to the target relation schema.
+			 */
+			if (change->data.tp.oldtuple != NULL)
+			{
+				HeapTuple sourceRelationOldTuple = change->data.tp.oldtuple;
+				HeapTuple targetRelationOldTuple = GetTupleForTargetSchemaForCdc(
+					sourceRelationOldTuple,
+					sourceRelationDesc,
+					targetRelationDesc);
+
+				change->data.tp.oldtuple = targetRelationOldTuple;
+			}
+			break;
+		}
+
+		case REORDER_BUFFER_CHANGE_DELETE:
+		{
+			/* For delete action, only old tuple should be translated*/
+			HeapTuple sourceRelationOldTuple = change->data.tp.oldtuple;
+			HeapTuple targetRelationOldTuple = GetTupleForTargetSchemaForCdc(
+				sourceRelationOldTuple,
+				sourceRelationDesc,
+				targetRelationDesc);
+
+			change->data.tp.oldtuple = targetRelationOldTuple;
+			break;
+		}
+
+		default:
+		{
+			/* Do nothing for other action types. */
+			break;
+		}
+	}
+#else
+
+	/* Check the ReorderBufferChange's action type and handle them accordingly.*/
+	switch (change->action)
+	{
+		case REORDER_BUFFER_CHANGE_INSERT:
+		{
+			/* For insert action, only new tuple should always be translated*/
+			HeapTuple sourceRelationNewTuple = &(change->data.tp.newtuple->tuple);
+			HeapTuple targetRelationNewTuple = GetTupleForTargetSchemaForCdc(
+				sourceRelationNewTuple, sourceRelationDesc, targetRelationDesc);
+			change->data.tp.newtuple->tuple = *targetRelationNewTuple;
+			break;
+		}
+
+		/*
+		 * For update changes both old and new tuples need to be translated for target relation
+		 * if the REPLICA IDENTITY is set to FULL. Otherwise, only the new tuple needs to be
+		 * translated for target relation.
+		 */
+		case REORDER_BUFFER_CHANGE_UPDATE:
+		{
+			/* For update action, new tuple should always be translated*/
+			/* Get the new tuple from the ReorderBufferChange, and translate it to target relation. */
+			HeapTuple sourceRelationNewTuple = &(change->data.tp.newtuple->tuple);
+			HeapTuple targetRelationNewTuple = GetTupleForTargetSchemaForCdc(
+				sourceRelationNewTuple, sourceRelationDesc, targetRelationDesc);
+			change->data.tp.newtuple->tuple = *targetRelationNewTuple;
+
+			/*
+			 * Format oldtuple according to the target relation. If the column values of replica
+			 * identiy change, then the old tuple is non-null and needs to be formatted according
+			 * to the target relation schema.
+			 */
+			if (change->data.tp.oldtuple != NULL)
+			{
+				HeapTuple sourceRelationOldTuple = &(change->data.tp.oldtuple->tuple);
+				HeapTuple targetRelationOldTuple = GetTupleForTargetSchemaForCdc(
+					sourceRelationOldTuple,
+					sourceRelationDesc,
+					targetRelationDesc);
+
+				change->data.tp.oldtuple->tuple = *targetRelationOldTuple;
+			}
+			break;
+		}
+
+		case REORDER_BUFFER_CHANGE_DELETE:
+		{
+			/* For delete action, only old tuple should be translated*/
+			HeapTuple sourceRelationOldTuple = &(change->data.tp.oldtuple->tuple);
+			HeapTuple targetRelationOldTuple = GetTupleForTargetSchemaForCdc(
+				sourceRelationOldTuple,
+				sourceRelationDesc,
+				targetRelationDesc);
+
+			change->data.tp.oldtuple->tuple = *targetRelationOldTuple;
+			break;
+		}
+
+		default:
+		{
+			/* Do nothing for other action types. */
+			break;
+		}
+	}
+#endif
+}

src/backend/distributed/cdc/cdc_decoder_utils.c

@@ -0,0 +1,447 @@
+/*-------------------------------------------------------------------------
+ *
+ * cdc_decoder_utils.c
+ *		CDC Decoder plugin utility functions for Citus
+ *
+ * Copyright (c) Citus Data, Inc.
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "cdc_decoder_utils.h"
+#include "fmgr.h"
+#include "miscadmin.h"
+
+#include "access/genam.h"
+#include "access/heapam.h"
+#include "catalog/pg_namespace.h"
+#include "commands/extension.h"
+#include "common/hashfn.h"
+#include "common/string.h"
+#include "utils/fmgroids.h"
+#include "utils/lsyscache.h"
+#include "utils/typcache.h"
+
+#include "distributed/pg_dist_partition.h"
+#include "distributed/pg_dist_shard.h"
+#include "distributed/relay_utility.h"
+
+static int32 LocalGroupId = -1;
+static Oid PgDistLocalGroupRelationId = InvalidOid;
+static Oid PgDistShardRelationId = InvalidOid;
+static Oid PgDistShardShardidIndexId = InvalidOid;
+static Oid PgDistPartitionRelationId = InvalidOid;
+static Oid PgDistPartitionLogicalrelidIndexId = InvalidOid;
+static bool IsCitusExtensionLoaded = false;
+
+#define COORDINATOR_GROUP_ID 0
+#define InvalidRepOriginId 0
+#define Anum_pg_dist_local_groupid 1
+#define GROUP_ID_UPGRADING -2
+
+
+static Oid DistLocalGroupIdRelationId(void);
+static int32 CdcGetLocalGroupId(void);
+static HeapTuple CdcPgDistPartitionTupleViaCatalog(Oid relationId);
+
+/*
+ * DistLocalGroupIdRelationId returns the relation id of the pg_dist_local_group
+ */
+static Oid
+DistLocalGroupIdRelationId(void)
+{
+	if (PgDistLocalGroupRelationId == InvalidOid)
+	{
+		PgDistLocalGroupRelationId = get_relname_relid("pg_dist_local_group",
+													   PG_CATALOG_NAMESPACE);
+	}
+	return PgDistLocalGroupRelationId;
+}
+
+
+/*
+ * DistShardRelationId returns the relation id of the pg_dist_shard
+ */
+static Oid
+DistShardRelationId(void)
+{
+	if (PgDistShardRelationId == InvalidOid)
+	{
+		PgDistShardRelationId = get_relname_relid("pg_dist_shard", PG_CATALOG_NAMESPACE);
+	}
+	return PgDistShardRelationId;
+}
+
+
+/*
+ * DistShardShardidIndexId returns the relation id of the pg_dist_shard_shardid_index
+ */
+static Oid
+DistShardShardidIndexId(void)
+{
+	if (PgDistShardShardidIndexId == InvalidOid)
+	{
+		PgDistShardShardidIndexId = get_relname_relid("pg_dist_shard_shardid_index",
+													  PG_CATALOG_NAMESPACE);
+	}
+	return PgDistShardShardidIndexId;
+}
+
+
+/*
+ * DistPartitionRelationId returns the relation id of the pg_dist_partition
+ */
+static Oid
+DistPartitionRelationId(void)
+{
+	if (PgDistPartitionRelationId == InvalidOid)
+	{
+		PgDistPartitionRelationId = get_relname_relid("pg_dist_partition",
+													  PG_CATALOG_NAMESPACE);
+	}
+	return PgDistPartitionRelationId;
+}
+
+
+static Oid
+DistPartitionLogicalRelidIndexId(void)
+{
+	if (PgDistPartitionLogicalrelidIndexId == InvalidOid)
+	{
+		PgDistPartitionLogicalrelidIndexId = get_relname_relid(
+			"pg_dist_partition_logicalrelid_index", PG_CATALOG_NAMESPACE);
+	}
+	return PgDistPartitionLogicalrelidIndexId;
+}
+
+
+/*
+ * CdcIsCoordinator function returns true if this node is identified as the
+ * schema/coordinator/master node of the cluster.
+ */
+bool
+CdcIsCoordinator(void)
+{
+	return (CdcGetLocalGroupId() == COORDINATOR_GROUP_ID);
+}
+
+
+/*
+ * CdcCitusHasBeenLoaded function returns true if the citus extension has been loaded.
+ */
+bool
+CdcCitusHasBeenLoaded()
+{
+	if (!IsCitusExtensionLoaded)
+	{
+		IsCitusExtensionLoaded = (get_extension_oid("citus", true) != InvalidOid);
+	}
+
+	return IsCitusExtensionLoaded;
+}
+
+
+/*
+ * ExtractShardIdFromTableName tries to extract shard id from the given table name,
+ * and returns the shard id if table name is formatted as shard name.
+ * Else, the function returns INVALID_SHARD_ID.
+ */
+uint64
+CdcExtractShardIdFromTableName(const char *tableName, bool missingOk)
+{
+	char *shardIdStringEnd = NULL;
+
+	/* find the last underscore and increment for shardId string */
+	char *shardIdString = strrchr(tableName, SHARD_NAME_SEPARATOR);
+	if (shardIdString == NULL && !missingOk)
+	{
+		ereport(ERROR, (errmsg("could not extract shardId from table name \"%s\"",
+							   tableName)));
+	}
+	else if (shardIdString == NULL && missingOk)
+	{
+		return INVALID_SHARD_ID;
+	}
+
+	shardIdString++;
+
+	errno = 0;
+	uint64 shardId = strtoull(shardIdString, &shardIdStringEnd, 0);
+
+	if (errno != 0 || (*shardIdStringEnd != '\0'))
+	{
+		if (!missingOk)
+		{
+			ereport(ERROR, (errmsg("could not extract shardId from table name \"%s\"",
+								   tableName)));
+		}
+		else
+		{
+			return INVALID_SHARD_ID;
+		}
+	}
+
+	return shardId;
+}
+
+
+/*
+ * CdcGetLocalGroupId returns the group identifier of the local node. The
+ * function assumes that pg_dist_local_group has exactly one row and has at
+ * least one column. Otherwise, the function errors out.
+ */
+static int32
+CdcGetLocalGroupId(void)
+{
+	ScanKeyData scanKey[1];
+	int scanKeyCount = 0;
+	int32 groupId = 0;
+
+	/*
+	 * Already set the group id, no need to read the heap again.
+	 */
+	if (LocalGroupId != -1)
+	{
+		return LocalGroupId;
+	}
+
+	Oid localGroupTableOid = DistLocalGroupIdRelationId();
+	if (localGroupTableOid == InvalidOid)
+	{
+		return 0;
+	}
+
+	Relation pgDistLocalGroupId = table_open(localGroupTableOid, AccessShareLock);
+
+	SysScanDesc scanDescriptor = systable_beginscan(pgDistLocalGroupId,
+													InvalidOid, false,
+													NULL, scanKeyCount, scanKey);
+
+	TupleDesc tupleDescriptor = RelationGetDescr(pgDistLocalGroupId);
+
+	HeapTuple heapTuple = systable_getnext(scanDescriptor);
+
+	if (HeapTupleIsValid(heapTuple))
+	{
+		bool isNull = false;
+		Datum groupIdDatum = heap_getattr(heapTuple,
+										  Anum_pg_dist_local_groupid,
+										  tupleDescriptor, &isNull);
+
+		groupId = DatumGetInt32(groupIdDatum);
+
+		/* set the local cache variable */
+		LocalGroupId = groupId;
+	}
+	else
+	{
+		/*
+		 * Upgrade is happening. When upgrading postgres, pg_dist_local_group is
+		 * temporarily empty before citus_finish_pg_upgrade() finishes execution.
+		 */
+		groupId = GROUP_ID_UPGRADING;
+	}
+
+	systable_endscan(scanDescriptor);
+	table_close(pgDistLocalGroupId, AccessShareLock);
+
+	return groupId;
+}
+
+
+/*
+ * CdcLookupShardRelationFromCatalog returns the logical relation oid a shard belongs to.
+ *
+ * Errors out if the shardId does not exist and missingOk is false.
+ * Returns InvalidOid if the shardId does not exist and missingOk is true.
+ */
+Oid
+CdcLookupShardRelationFromCatalog(int64 shardId, bool missingOk)
+{
+	ScanKeyData scanKey[1];
+	int scanKeyCount = 1;
+	Form_pg_dist_shard shardForm = NULL;
+	Relation pgDistShard = table_open(DistShardRelationId(), AccessShareLock);
+	Oid relationId = InvalidOid;
+
+	ScanKeyInit(&scanKey[0], Anum_pg_dist_shard_shardid,
+				BTEqualStrategyNumber, F_INT8EQ, Int64GetDatum(shardId));
+
+	SysScanDesc scanDescriptor = systable_beginscan(pgDistShard,
+													DistShardShardidIndexId(), true,
+													NULL, scanKeyCount, scanKey);
+
+	HeapTuple heapTuple = systable_getnext(scanDescriptor);
+	if (!HeapTupleIsValid(heapTuple) && !missingOk)
+	{
+		ereport(ERROR, (errmsg("could not find valid entry for shard "
+							   UINT64_FORMAT, shardId)));
+	}
+
+	if (!HeapTupleIsValid(heapTuple))
+	{
+		relationId = InvalidOid;
+	}
+	else
+	{
+		shardForm = (Form_pg_dist_shard) GETSTRUCT(heapTuple);
+		relationId = shardForm->logicalrelid;
+	}
+
+	systable_endscan(scanDescriptor);
+	table_close(pgDistShard, NoLock);
+
+	return relationId;
+}
+
+
+/*
+ * CdcPgDistPartitionTupleViaCatalog is a helper function that searches
+ * pg_dist_partition for the given relationId. The caller is responsible
+ * for ensuring that the returned heap tuple is valid before accessing
+ * its fields.
+ */
+static HeapTuple
+CdcPgDistPartitionTupleViaCatalog(Oid relationId)
+{
+	const int scanKeyCount = 1;
+	ScanKeyData scanKey[1];
+	bool indexOK = true;
+
+	Relation pgDistPartition = table_open(DistPartitionRelationId(), AccessShareLock);
+
+	ScanKeyInit(&scanKey[0], Anum_pg_dist_partition_logicalrelid,
+				BTEqualStrategyNumber, F_OIDEQ, ObjectIdGetDatum(relationId));
+
+	SysScanDesc scanDescriptor = systable_beginscan(pgDistPartition,
+													DistPartitionLogicalRelidIndexId(),
+													indexOK, NULL, scanKeyCount, scanKey);
+
+	HeapTuple partitionTuple = systable_getnext(scanDescriptor);
+
+	if (HeapTupleIsValid(partitionTuple))
+	{
+		/* callers should have the tuple in their memory contexts */
+		partitionTuple = heap_copytuple(partitionTuple);
+	}
+
+	systable_endscan(scanDescriptor);
+	table_close(pgDistPartition, AccessShareLock);
+
+	return partitionTuple;
+}
+
+
+/*
+ * CdcIsReferenceTableViaCatalog gets a relationId and returns true if the relation
+ * is a reference table and false otherwise.
+ */
+char
+CdcIsReferenceTableViaCatalog(Oid relationId)
+{
+	HeapTuple partitionTuple = CdcPgDistPartitionTupleViaCatalog(relationId);
+	if (!HeapTupleIsValid(partitionTuple))
+	{
+		return false;
+	}
+
+	Datum datumArray[Natts_pg_dist_partition];
+	bool isNullArray[Natts_pg_dist_partition];
+
+	Relation pgDistPartition = table_open(DistPartitionRelationId(), AccessShareLock);
+
+	TupleDesc tupleDescriptor = RelationGetDescr(pgDistPartition);
+	heap_deform_tuple(partitionTuple, tupleDescriptor, datumArray, isNullArray);
+
+	if (isNullArray[Anum_pg_dist_partition_partmethod - 1] ||
+		isNullArray[Anum_pg_dist_partition_repmodel - 1])
+	{
+		/*
+		 * partition method and replication model cannot be NULL,
+		 * still let's make sure
+		 */
+		heap_freetuple(partitionTuple);
+		table_close(pgDistPartition, NoLock);
+		return false;
+	}
+
+	Datum partitionMethodDatum = datumArray[Anum_pg_dist_partition_partmethod - 1];
+	char partitionMethodChar = DatumGetChar(partitionMethodDatum);
+
+	Datum replicationModelDatum = datumArray[Anum_pg_dist_partition_repmodel - 1];
+	char replicationModelChar = DatumGetChar(replicationModelDatum);
+
+	heap_freetuple(partitionTuple);
+	table_close(pgDistPartition, NoLock);
+
+	/*
+	 * A table is a reference table when its partition method is 'none'
+	 * and replication model is 'two phase commit'
+	 */
+	return partitionMethodChar == DISTRIBUTE_BY_NONE &&
+		   replicationModelChar == REPLICATION_MODEL_2PC;
+}
+
+
+/*
+ * RemoveCitusDecodersFromPaths removes a path ending in citus_decoders
+ * from the given input paths.
+ */
+char *
+RemoveCitusDecodersFromPaths(char *paths)
+{
+	if (strlen(paths) == 0)
+	{
+		/* dynamic_library_path is empty */
+		return paths;
+	}
+
+	StringInfo newPaths = makeStringInfo();
+
+	char *remainingPaths = paths;
+
+	for (;;)
+	{
+		int pathLength = 0;
+
+		char *pathStart = first_path_var_separator(remainingPaths);
+		if (pathStart == remainingPaths)
+		{
+			/*
+			 * This will error out in find_in_dynamic_libpath, return
+			 * original value here.
+			 */
+			return paths;
+		}
+		else if (pathStart == NULL)
+		{
+			/* final path */
+			pathLength = strlen(remainingPaths);
+		}
+		else
+		{
+			/* more paths remaining */
+			pathLength = pathStart - remainingPaths;
+		}
+
+		char *currentPath = palloc(pathLength + 1);
+		strlcpy(currentPath, remainingPaths, pathLength + 1);
+		canonicalize_path(currentPath);
+
+		if (!pg_str_endswith(currentPath, "/citus_decoders"))
+		{
+			appendStringInfo(newPaths, "%s%s", newPaths->len > 0 ? ":" : "", currentPath);
+		}
+
+		if (remainingPaths[pathLength] == '\0')
+		{
+			/* end of string */
+			break;
+		}
+
+		remainingPaths += pathLength + 1;
+	}
+
+	return newPaths->data;
+}

src/backend/distributed/cdc/cdc_decoder_utils.h

@@ -0,0 +1,36 @@
+/*-------------------------------------------------------------------------
+ *
+ * cdc_decoder_utils.h
+ *	  Utility functions and declerations for cdc decoder.
+ *
+ * Copyright (c) Citus Data, Inc.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#ifndef CITUS_CDC_DECODER_H
+#define CITUS_CDC_DECODER_H
+
+#include "postgres.h"
+
+#include "c.h"
+#include "fmgr.h"
+
+#include "replication/logical.h"
+
+#define InvalidRepOriginId 0
+#define INVALID_SHARD_ID 0
+
+bool CdcIsCoordinator(void);
+
+uint64 CdcExtractShardIdFromTableName(const char *tableName, bool missingOk);
+
+Oid CdcLookupShardRelationFromCatalog(int64 shardId, bool missingOk);
+
+char CdcIsReferenceTableViaCatalog(Oid relationId);
+
+bool CdcCitusHasBeenLoaded(void);
+
+char * RemoveCitusDecodersFromPaths(char *paths);
+
+#endif   /* CITUS_CDC_DECODER_UTILS_H */

src/backend/distributed/citus.control

@@ -1,6 +1,6 @@
 # Citus extension
 comment = 'Citus distributed database'
-default_version = '11.1-1'
+default_version = '13.2-1'
 module_pathname = '$libdir/citus'
 relocatable = false
 schema = pg_catalog

src/backend/distributed/clock/README.md

@@ -0,0 +1,244 @@
+# Cluster Clock
+### Requirement:
+Many distributed applications need to track the changes in the same order as they are applied on the database. The changes can be to databases or objects within them, either within a single node or across the sharded cluster.
+### Definitions
+**Total ordering** - Every pair of change events can be placed in some order.
+**Causal ordering** - Only events that are causally related (an event A caused an event B) can be ordered i.e., it's only a partial order - sometimes events happen independently with no possible causal relationship, such events are treated to concurrent.
+**Sequential consistency** - All writes must be seen in the same order by all processes.
+**Causal consistency** - Causally related writes must be seen in the same order.
+
+Transactions on a single node system naturally provide a total and sequential ordering guarantees for client read and write operations as all operations are routed to the same node, but there are challenges for a multi node distributed system, such as, Citus.
+
+One possible way to totally order all the changes in the system is to timestamp all the events with a global physical clock or a centralized logical clock. Thus, observing the events in the increasing order of the timestamp will give the total ordering of events. For both the performance and cost reasons such solutions are impractical. In the absence of total ordering, a little weaker ordering is the **causal order**.
+
+Causal order is defined as a model that preserves a partial order of events in a distributed system. If an event
+
+1.  A causes another event B, every other process in the system observes the event A before observing event B.
+2.  Causal order is transitive: if A causes B, and B causes C, then A causes C.
+3.  Non causally ordered events are treated as concurrent.
+
+Causal consistency is a weak form of consistency that preserves the order of causally related operations. The causal consistency model can be refined into four session guarantees.
+
+1.  Read Your Writes: If a process performs a write, the same process later observes the result of its write.
+6.  Monotonic Reads: The set of writes observed (read) by a process is guaranteed to be monotonically increasing.
+7.  Writes Follow Reads: If some process performs a read followed by a write, and another process observes the result of the write, then it can also observe the read.
+8.  Monotonic Writes: If some process performs a write, followed sometime later by another write, other processes will observe them in the same order.
+
+### Hybrid Logical Clock (HLC)
+HLC provides a way to get the causality relationship like logical clocks. It can also be used for backup/recovery too as the logical clock value is maintained close to the wall clock time. HLC consists of
+ LC - Logical clock
+ C - Counter
+
+Clock components - Unsigned 64 bit <LC, C>
+Epoch Milliseconds ( LC ) | Logical counter ( C )|
+|--|--|
+| 42 bits | 22 bits |
+
+2^42 milliseconds - 4398046511104 milliseconds, which is ~139 years.
+
+2^22 ticks - maximum of four million operations per millisecond.
+
+### UDFs
+A new UDF `citus_get_cluster_clock`() that returns a monotonically increasing logical clock. Clock guarantees to never go back in value after restarts and makes best attempt to keep the value close to UNIX epoch time in milliseconds.
+
+A new UDF `citus_get_transaction_clock`(), when called by the user, returns the logical causal clock timestamp current transaction,
+Internally, this is the maximum clock among all transaction nodes, and
+all nodes move to the new clock.
+
+### GUC
+A new GUC parameter, "**citus.enable_cluster_clock**", If clocks go bad for any reason, this serves as a safety valve to avoid the need to change the application and (re)deploy it.
+
+### Sequence
+In Unix, though rare, there is a possibility of clock drifting backwards (or
+forward) after a restart. In such rare scenarios, we might end up with a logical clock value less than the previously used value, this violates the fundamental requirement of monotonically increasing clock. To avoid such disasters, every logical clock tick is persisted using sequences (non-transactional). After a restart, the persisted sequence value is read and clock starts from that value, which will ensure that system starts the clock from where we left off.
+
+### Psuedo code
+    WC - Current Wall Clock in milliseconds
+    HLC - Current Hybrid Logical Clock in shared
+          memory
+    MAX_COUNTER - Four million
+
+    /* Tick the clock by 1 */
+    IncrementClusterClock()
+    {
+       /* It's the counter that always ticks, once it reaches
+          the maximum, reset the counter to 1 and increment
+          the logical clock. */
+
+        if (HLC.C == MAX_COUNTER)
+        {
+                HLC.LC++;
+                HLC.C = 0;
+                return;
+        }
+        HLC.C++;
+    }
+
+    /* Tick for each event, must increase monotonically */
+    GetNextNodeClockValue()
+    {
+	    IncrementClusterClock(HLC);
+
+	    /* From the incremented clock and current wall clock,
+	       pick which ever is highest */
+	    NextClock = MAX(HLC, WC);
+
+	    /* Save the NextClock value in both the shared memory
+	       and sequence */
+	    HLC = NextClock;
+	    SETVAL(pg_dist_clock_logical_seq, HLC);
+	}
+
+	/* Returns true if the clock1 is after clock2 */
+    IsEventAfter(HLC1, HLC2)
+    {
+	    IF (HLC1.LC != HLC2.LC)
+		    return (HLC1.LC > HLC2.LC);
+		ELSE
+			return (HLC1.C > HLC2.C);
+    }
+
+    /* Simply returns the highest node clock value among all
+       nodes */
+    GetHighestClockInTransaction()
+    {
+	    For each node
+	    {
+		    NodeClock[N] = GetNextNodeClockValue();
+		}
+
+		/* Return the highest clock value of all the nodes */
+		return MAX(NodeClock[N]);
+    }
+
+    /* Adjust the local shared memory clock to the received
+       value (RHLC) from the remote node */
+	AdjustClock(RHLC)
+	{
+		/* local clock is ahead or equal, do nothing */
+        IF (HLC >= RHLC)
+        {
+          return;
+        }
+        /* Save the remote clockvalue in both the shared
+           memory and sequence */
+	    HLC = RHLC;
+	    SETVAL(pg_dist_clock_logical_seq, HLC);
+	}
+
+    /* All the nodes will adjust their clocks to the highest
+       of the newly negotiated clock */
+    AdjustClocksToTransactionHighest(HLC)
+    {
+	    For each node
+	    {
+		    SendCommand ("AdjustClock(HLC)");
+		}
+    }
+
+    /* When citus_get_transaction_clock() UDF is invoked */
+    PrepareAndSetTransactionClock()
+    {
+	    /* Pick the highest logical clock value among all
+	       transaction-nodes */
+        txnCLock = GetHighestClockInTransaction()
+
+        /* Adjust all the nodes with the new clock value */
+        AdjustClocksToTransactionHighest(txnCLock )
+
+        return txnClock;
+    }
+
+    /* Initialize the clock value to the highest clock
+       persisted in sequence */
+    InitClockAtBoot()
+    {
+	    /* Start with the current wall clock */
+	    HLC = WC;
+
+	    IF (SEQUENCE == 1)
+		/* clock never ticked on this node, start with the
+		   wall clock. */
+				return;
+		/* get the most recent clock ever used from disk */
+		persistedClock =
+		    NEXT_VAL(pg_dist_clock_logical_seq...)
+		    /* Start the clock with persisted value */
+		    AdjustLocalClock(persistedClock);
+	    }
+    }
+
+ #### Usage
+**Step 1**
+In the application, track every change of a transaction along with the unique transaction ID by calling UDF
+`get_current_transaction_id`()
+
+    INSERT INTO track_table
+    SET TransactionId =
+    get_current_transaction_id(),
+    operation = <insert/update/delete>,
+    row_key = <>,
+    ....;
+
+**Step 2**
+As the transaction is about to end, and before the COMMIT, capture the causal clock timestamp along with the transaction ID in a table
+
+    INSERT INTO transaction_commit_clock
+    (TransactionId, CommitClock, timestamp)
+    SELECT
+            citus_get_transaction_clock(),
+            get_current_transaction_id(),
+            now()
+
+**Step 3**
+How to get all the events in the causal order?
+
+    SELECT tt.row_key, tt.operation
+    FROM track_table tt,
+         transaction_commit_clock cc
+    WHERE tt.TransactionId = cc.TransactionId
+    ORDER BY cc.CommitClock
+
+Events for an object
+
+    SELECT tt.row_key, tt.operation
+    FROM track_table tt,
+         transaction_commit_clock cc
+    WHERE tt.TransactionId = cc.TransactionId
+    and row_key = $1 ORDER BY cc.CommitClock
+
+Events in the last one hour
+
+    SELECT tt.row_key, tt.operation
+    FROM track_table tt,
+         transaction_commit_clockcc
+    WHERE cc.timestamp >= now() - interval '1 hour'
+    and tt.TransactionId = cc.TransactionId
+
+**Note**: In Citus we use 2PC, if any node goes down after the PREPARE and before the COMMIT, we might have changes partially committed. Citus tracks such transactions in **pg_dist_transaction** and eventually will be committed when the node becomes healthy, but when we track change-data from committed transactions of **transaction_commit_clock** we will miss the changes from a bad node.
+
+To address this issue, proposal is to have a new UDF #TBD, that freezes
+the clock and ensures that all the 2PCs are fully complete
+(i.e., **pg_dist_transaction** should be empty) and return the highest
+clock used. All transactions in `transaction_commit_clock` with
+timestamp below this returned clock are visible to the application. The
+exact nuances, such as frequency of calling such UDF, are still TBD.
+Caveat is, if the node and the 2PC takes long to fully recover, the
+visibility of the committed transactions might stall.
+
+### Catalog pruning
+The data in **transaction_commit_clock** should be ephemeral data i.e., eventually rows have to automatically be deleted. Users can install a pg_cron job to prune the catalog regularly.
+
+    delete from transaction_commit_clock
+        where timestamp < now() - interval '7 days'
+
+### Limitations of Citus
+Using this transaction commit clock ordering to build a secondary, that's a mirror copy of the original, may not be feasible at this time for the following reasons.
+
+Given that there is no well-defined order between concurrent distributed transactions in Citus, we cannot retroactively apply a transaction-order that leads to an exact replica of the primary unless we preserve the original object-level ordering as it happened on individual nodes.
+
+For instance, if a multi-shard insert (transaction A) happens concurrently with a multi-shard update (transaction B) and the WHERE clause of the update matches inserted rows in multiple shards, we could have a scenario in which only a subset of the inserted rows gets updated. Effectively, transaction A might happen before transaction B on node 1, while transaction B happens before transaction A on node 2. While unfortunate, we cannot simply claim changes made by transaction A happened first based on commit timestamps, because that would lead us reorder changes to the same object ID on node 2, which might lead to a different outcome when replayed.
+
+In such scenario, even if we use causal commit clock to order changes. It is essential that the order of modifications to an object matches the original order. Otherwise, you could have above scenarios where an insert happens before an update in the primary cluster, but the update happens before the insert. Replaying the changes would then lead to a different database.
+
+In absence of a coherent transaction-ordering semantics in distributed cluster, best we can do is ensure that changes to the same object are in the correct order and ensure exactly once delivery (correct pagination).

src/backend/distributed/clock/causal_clock.c

@@ -0,0 +1,622 @@
+/*-------------------------------------------------------------------------
+ * causal_clock.c
+ *
+ * Core function defintions to implement hybrid logical clock.
+ *
+ * Copyright (c) Citus Data, Inc.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#include <sys/time.h>
+
+#include "postgres.h"
+
+#include "fmgr.h"
+#include "funcapi.h"
+#include "libpq-fe.h"
+#include "miscadmin.h"
+
+#include "catalog/namespace.h"
+#include "commands/extension.h"
+#include "commands/sequence.h"
+#include "executor/spi.h"
+#include "nodes/pg_list.h"
+#include "postmaster/postmaster.h"
+#include "storage/ipc.h"
+#include "storage/lwlock.h"
+#include "storage/s_lock.h"
+#include "storage/shmem.h"
+#include "storage/spin.h"
+#include "utils/builtins.h"
+#include "utils/datum.h"
+#include "utils/numeric.h"
+#include "utils/typcache.h"
+
+#include "distributed/causal_clock.h"
+#include "distributed/citus_safe_lib.h"
+#include "distributed/coordinator_protocol.h"
+#include "distributed/listutils.h"
+#include "distributed/local_executor.h"
+#include "distributed/lock_graph.h"
+#include "distributed/metadata_cache.h"
+#include "distributed/placement_connection.h"
+#include "distributed/remote_commands.h"
+
+#define SAVE_AND_PERSIST(c) \
+	do { \
+		Oid savedUserId = InvalidOid; \
+		int savedSecurityContext = 0; \
+		LogicalClockShmem->clusterClockValue = *(c); \
+		GetUserIdAndSecContext(&savedUserId, &savedSecurityContext); \
+		SetUserIdAndSecContext(CitusExtensionOwner(), SECURITY_LOCAL_USERID_CHANGE); \
+		DirectFunctionCall2(setval_oid, \
+							ObjectIdGetDatum(DistClockLogicalSequenceId()), \
+							Int64GetDatum((c)->logical)); \
+		SetUserIdAndSecContext(savedUserId, savedSecurityContext); \
+	} while (0)
+
+PG_FUNCTION_INFO_V1(citus_get_node_clock);
+PG_FUNCTION_INFO_V1(citus_internal_adjust_local_clock_to_remote);
+PG_FUNCTION_INFO_V1(citus_is_clock_after);
+PG_FUNCTION_INFO_V1(citus_get_transaction_clock);
+
+/*
+ * Current state of the logical clock
+ */
+typedef enum ClockState
+{
+	CLOCKSTATE_INITIALIZED,
+	CLOCKSTATE_UNINITIALIZED
+} ClockState;
+
+/*
+ * Holds the cluster clock variables in shared memory.
+ */
+typedef struct LogicalClockShmemData
+{
+	NamedLWLockTranche namedLockTranche;
+	LWLock clockLock;
+
+	/* Current logical clock value of this node */
+	ClusterClock clusterClockValue;
+
+	/* Tracks initialization at boot */
+	ClockState clockInitialized;
+} LogicalClockShmemData;
+
+
+static shmem_startup_hook_type prev_shmem_startup_hook = NULL;
+static LogicalClockShmemData *LogicalClockShmem = NULL;
+static void AdjustLocalClock(ClusterClock *remoteClock);
+static void GetNextNodeClockValue(ClusterClock *nextClusterClockValue);
+static ClusterClock * GetHighestClockInTransaction(List *nodeConnectionList);
+static void AdjustClocksToTransactionHighest(List *nodeConnectionList,
+											 ClusterClock *transactionClockValue);
+static void InitClockAtFirstUse(void);
+static void IncrementClusterClock(ClusterClock *clusterClock);
+static ClusterClock * LargerClock(ClusterClock *clock1, ClusterClock *clock2);
+static ClusterClock * PrepareAndSetTransactionClock(void);
+bool EnableClusterClock = true;
+
+
+/*
+ * GetEpochTimeAsClock returns the epoch value milliseconds used as logical
+ * value in ClusterClock.
+ */
+ClusterClock *
+GetEpochTimeAsClock(void)
+{
+	struct timeval tp = { 0 };
+
+	gettimeofday(&tp, NULL);
+
+	uint64 result = (uint64) (tp.tv_sec) * 1000;
+	result = result + (uint64) (tp.tv_usec) / 1000;
+
+	ClusterClock *epochClock = (ClusterClock *) palloc(sizeof(ClusterClock));
+	epochClock->logical = result;
+	epochClock->counter = 0;
+
+	return epochClock;
+}
+
+
+/*
+ * LogicalClockShmemSize returns the size that should be allocated
+ * in the shared memory for logical clock management.
+ */
+size_t
+LogicalClockShmemSize(void)
+{
+	Size size = 0;
+
+	size = add_size(size, sizeof(LogicalClockShmemData));
+
+	return size;
+}
+
+
+/*
+ * InitializeClusterClockMem reserves shared-memory space needed to
+ * store LogicalClockShmemData, and sets the hook for initialization
+ * of the same.
+ */
+void
+InitializeClusterClockMem(void)
+{
+	prev_shmem_startup_hook = shmem_startup_hook;
+	shmem_startup_hook = LogicalClockShmemInit;
+}
+
+
+/*
+ * LogicalClockShmemInit Allocates and initializes shared memory for
+ * cluster clock related variables.
+ */
+void
+LogicalClockShmemInit(void)
+{
+	bool alreadyInitialized = false;
+
+	LWLockAcquire(AddinShmemInitLock, LW_EXCLUSIVE);
+
+	LogicalClockShmem = (LogicalClockShmemData *)
+						ShmemInitStruct("Logical Clock Shmem",
+										LogicalClockShmemSize(),
+										&alreadyInitialized);
+
+	if (!alreadyInitialized)
+	{
+		/* A zero value indicates that the clock is not adjusted yet */
+		memset(&LogicalClockShmem->clusterClockValue, 0, sizeof(ClusterClock));
+
+		LogicalClockShmem->namedLockTranche.trancheName = "Cluster Clock Setup Tranche";
+		LogicalClockShmem->namedLockTranche.trancheId = LWLockNewTrancheId();
+		LWLockRegisterTranche(LogicalClockShmem->namedLockTranche.trancheId,
+							  LogicalClockShmem->namedLockTranche.trancheName);
+		LWLockInitialize(&LogicalClockShmem->clockLock,
+						 LogicalClockShmem->namedLockTranche.trancheId);
+
+		LogicalClockShmem->clockInitialized = CLOCKSTATE_UNINITIALIZED;
+	}
+
+	LWLockRelease(AddinShmemInitLock);
+
+	if (prev_shmem_startup_hook != NULL)
+	{
+		prev_shmem_startup_hook();
+	}
+}
+
+
+/*
+ * IncrementClusterClock increments the ClusterClock by 1.
+ */
+static void
+IncrementClusterClock(ClusterClock *clusterClock)
+{
+	/*
+	 * It's the counter that always ticks, once it reaches the maximum, reset
+	 * the counter to 1 and increment the logical clock.
+	 */
+	if (clusterClock->counter == MAX_COUNTER)
+	{
+		clusterClock->logical++;
+		clusterClock->counter = 0;
+		return;
+	}
+
+	clusterClock->counter++;
+}
+
+
+/*
+ * LargerClock compares two ClusterClock(s) and returns pointer to the larger one.
+ * Note: If equal or one of the clock is NULL, non NULL clock is copied.
+ */
+static ClusterClock *
+LargerClock(ClusterClock *clock1, ClusterClock *clock2)
+{
+	/* Check if one of the paramater is NULL */
+	if (!clock1 || !clock2)
+	{
+		Assert(clock1 || clock2);
+		return (!clock1 ? clock2 : clock1);
+	}
+
+	if (cluster_clock_cmp_internal(clock1, clock2) > 0)
+	{
+		return clock1;
+	}
+	else
+	{
+		return clock2;
+	}
+}
+
+
+/*
+ * GetNextNodeClock implements the internal guts of the UDF citus_get_node_clock()
+ */
+static void
+GetNextNodeClockValue(ClusterClock *nextClusterClockValue)
+{
+	static bool isClockInitChecked = false; /* serves as a local cache */
+	ClusterClock *epochValue = GetEpochTimeAsClock();
+
+	/* If this backend already checked for initialization, skip it */
+	if (!isClockInitChecked)
+	{
+		InitClockAtFirstUse();
+
+		/* We reach here only if CLOCKSTATE_INITIALIZED, all other cases error out. */
+		isClockInitChecked = true;
+	}
+
+	LWLockAcquire(&LogicalClockShmem->clockLock, LW_EXCLUSIVE);
+
+	Assert(LogicalClockShmem->clockInitialized == CLOCKSTATE_INITIALIZED);
+
+	/* Tick the clock */
+	IncrementClusterClock(&LogicalClockShmem->clusterClockValue);
+
+	/* Pick the larger of the two, wallclock and logical clock. */
+	ClusterClock *clockValue = LargerClock(&LogicalClockShmem->clusterClockValue,
+										   epochValue);
+
+	/*
+	 * Save the returned value in both the shared memory and sequences.
+	 */
+	SAVE_AND_PERSIST(clockValue);
+
+	LWLockRelease(&LogicalClockShmem->clockLock);
+
+	/* Return the clock */
+	*nextClusterClockValue = *clockValue;
+}
+
+
+/*
+ * AdjustLocalClock Adjusts the local shared memory clock to the
+ * received value from the remote node.
+ */
+void
+AdjustLocalClock(ClusterClock *remoteClock)
+{
+	LWLockAcquire(&LogicalClockShmem->clockLock, LW_EXCLUSIVE);
+
+	ClusterClock *localClock = &LogicalClockShmem->clusterClockValue;
+
+	/* local clock is ahead or equal, do nothing */
+	if (cluster_clock_cmp_internal(localClock, remoteClock) >= 0)
+	{
+		LWLockRelease(&LogicalClockShmem->clockLock);
+		return;
+	}
+
+	SAVE_AND_PERSIST(remoteClock);
+
+	LWLockRelease(&LogicalClockShmem->clockLock);
+
+	ereport(DEBUG1, (errmsg("adjusted to remote clock: "
+							"<logical(%lu) counter(%u)>",
+							remoteClock->logical,
+							remoteClock->counter)));
+}
+
+
+/*
+ * GetHighestClockInTransaction takes the connection list of participating nodes in the
+ * current transaction and polls the logical clock value of all the nodes. Returns the
+ * highest logical clock value of all the nodes in the current distributed transaction,
+ * which may be used as commit order for individual objects in the transaction.
+ */
+static ClusterClock *
+GetHighestClockInTransaction(List *nodeConnectionList)
+{
+	MultiConnection *connection = NULL;
+
+	foreach_declared_ptr(connection, nodeConnectionList)
+	{
+		int querySent =
+			SendRemoteCommand(connection, "SELECT citus_get_node_clock();");
+
+		if (querySent == 0)
+		{
+			ReportConnectionError(connection, ERROR);
+		}
+	}
+
+	/* Check for the current node */
+	ClusterClock *globalClockValue = (ClusterClock *) palloc(sizeof(ClusterClock));
+
+	GetNextNodeClockValue(globalClockValue);
+
+	ereport(DEBUG1, (errmsg("node(%u) transaction clock %lu:%u",
+							PostPortNumber, globalClockValue->logical,
+							globalClockValue->counter)));
+
+	/* fetch the results and pick the highest clock value of all the nodes */
+	foreach_declared_ptr(connection, nodeConnectionList)
+	{
+		bool raiseInterrupts = true;
+
+		if (PQstatus(connection->pgConn) != CONNECTION_OK)
+		{
+			ereport(ERROR, (errmsg("connection to %s:%d failed when "
+								   "fetching logical clock value",
+								   connection->hostname, connection->port)));
+		}
+
+		PGresult *result = GetRemoteCommandResult(connection, raiseInterrupts);
+		if (!IsResponseOK(result))
+		{
+			ReportResultError(connection, result, ERROR);
+		}
+
+		ClusterClock *nodeClockValue = ParseClusterClockPGresult(result, 0, 0);
+
+		ereport(DEBUG1, (errmsg("node(%u) transaction clock %lu:%u",
+								connection->port, nodeClockValue->logical,
+								nodeClockValue->counter)));
+
+		globalClockValue = LargerClock(globalClockValue, nodeClockValue);
+
+		PQclear(result);
+		ForgetResults(connection);
+	}
+
+	ereport(DEBUG1, (errmsg("final global transaction clock %lu:%u",
+							globalClockValue->logical,
+							globalClockValue->counter)));
+	return globalClockValue;
+}
+
+
+/*
+ * AdjustClocksToTransactionHighest Sets the clock value of all the nodes, participated
+ * in the PREPARE of the transaction, to the highest clock value of all the nodes.
+ */
+static void
+AdjustClocksToTransactionHighest(List *nodeConnectionList,
+								 ClusterClock *transactionClockValue)
+{
+	StringInfo queryToSend = makeStringInfo();
+
+	/* Set the clock value on participating worker nodes */
+	appendStringInfo(queryToSend,
+					 "SELECT citus_internal.adjust_local_clock_to_remote"
+					 "('(%lu, %u)'::pg_catalog.cluster_clock);",
+					 transactionClockValue->logical, transactionClockValue->counter);
+
+	ExecuteRemoteCommandInConnectionList(nodeConnectionList, queryToSend->data);
+	AdjustLocalClock(transactionClockValue);
+}
+
+
+/*
+ * PrepareAndSetTransactionClock polls all the transaction-nodes for their respective clocks,
+ * picks the highest clock and returns it via UDF citus_get_transaction_clock. All the nodes
+ * will now move to this newly negotiated clock.
+ */
+static ClusterClock *
+PrepareAndSetTransactionClock(void)
+{
+	if (!EnableClusterClock)
+	{
+		/* citus.enable_cluster_clock is false */
+		ereport(WARNING, (errmsg("GUC enable_cluster_clock is off")));
+		return NULL;
+	}
+
+	dlist_iter iter;
+	List *transactionNodeList = NIL;
+	List *nodeList = NIL;
+
+	/* Prepare the connection list */
+	dlist_foreach(iter, &InProgressTransactions)
+	{
+		MultiConnection *connection = dlist_container(MultiConnection, transactionNode,
+													  iter.cur);
+		WorkerNode *workerNode = FindWorkerNode(connection->hostname, connection->port);
+		if (!workerNode)
+		{
+			ereport(WARNING, errmsg("Worker node is missing"));
+			continue;
+		}
+
+		/* Skip the node if we already in the list */
+		if (list_member_int(nodeList, workerNode->groupId))
+		{
+			continue;
+		}
+
+		RemoteTransaction *transaction = &connection->remoteTransaction;
+
+		Assert(transaction->transactionState != REMOTE_TRANS_NOT_STARTED);
+
+		/* Skip a transaction that failed */
+		if (transaction->transactionFailed)
+		{
+			continue;
+		}
+
+		nodeList = lappend_int(nodeList, workerNode->groupId);
+		transactionNodeList = lappend(transactionNodeList, connection);
+	}
+
+	/* Pick the highest logical clock value among all transaction-nodes */
+	ClusterClock *transactionClockValue =
+		GetHighestClockInTransaction(transactionNodeList);
+
+	/* Adjust all the nodes with the new clock value */
+	AdjustClocksToTransactionHighest(transactionNodeList, transactionClockValue);
+
+	return transactionClockValue;
+}
+
+
+/*
+ * InitClockAtFirstUse Initializes the shared memory clock value to the highest clock
+ * persisted. This will protect from any clock drifts.
+ */
+static void
+InitClockAtFirstUse(void)
+{
+	LWLockAcquire(&LogicalClockShmem->clockLock, LW_EXCLUSIVE);
+
+	/* Avoid repeated and parallel initialization */
+	if (LogicalClockShmem->clockInitialized == CLOCKSTATE_INITIALIZED)
+	{
+		LWLockRelease(&LogicalClockShmem->clockLock);
+		return;
+	}
+
+	if (DistClockLogicalSequenceId() == InvalidOid)
+	{
+		ereport(ERROR, (errmsg("Clock related sequence is missing")));
+	}
+
+	/* Start with the wall clock value */
+	ClusterClock *epochValue = GetEpochTimeAsClock();
+	LogicalClockShmem->clusterClockValue = *epochValue;
+
+	/*  Retrieve the highest clock value persisted in the sequence */
+	ClusterClock persistedMaxClock = { 0 };
+
+	/*
+	 * We will get one more than the persisted value, but that's harmless and
+	 * also very _crucial_ in below scenarios
+	 *
+	 * 1) As sequences are not transactional, this will protect us from crashes
+	 *    after the logical increment and before the counter increment.
+	 *
+	 * 2) If a clock drifts backwards, we should always start one clock above
+	 *    the previous value, though we are not persisting the counter as the
+	 *    logical value supersedes the counter, a simple increment of it will
+	 *    protect us.
+	 *
+	 * Note: The first (and every 32nd) call to nextval() consumes 32 values in the
+	 * WAL. This is an optimization that postgres does to only have to write a WAL
+	 * entry every 32 invocations. Normally this is harmless, however, if the database
+	 * gets in a crashloop it could outrun the wall clock, if the database crashes at
+	 * a higher rate than once every 32 seconds.
+	 *
+	 */
+	Oid saveUserId = InvalidOid;
+	int savedSecurityCtx = 0;
+
+	GetUserIdAndSecContext(&saveUserId, &savedSecurityCtx);
+	SetUserIdAndSecContext(CitusExtensionOwner(), SECURITY_LOCAL_USERID_CHANGE);
+
+	persistedMaxClock.logical =
+		DirectFunctionCall1(nextval_oid, ObjectIdGetDatum(DistClockLogicalSequenceId()));
+
+	SetUserIdAndSecContext(saveUserId, savedSecurityCtx);
+
+	/*
+	 * Sequence 1 indicates no prior clock timestamps on this server, retain
+	 * the wall clock i.e. no adjustment needed.
+	 */
+	if (persistedMaxClock.logical != 1)
+	{
+		ereport(DEBUG1, (errmsg("adjusting the clock with persisted value: "
+								"<logical(%lu) and counter(%u)>",
+								persistedMaxClock.logical,
+								persistedMaxClock.counter)));
+
+		/*
+		 * Adjust the local clock according to the most recent
+		 * clock stamp value persisted in the catalog.
+		 */
+		if (cluster_clock_cmp_internal(&persistedMaxClock, epochValue) > 0)
+		{
+			SAVE_AND_PERSIST(&persistedMaxClock);
+			ereport(NOTICE, (errmsg("clock drifted backwards, adjusted to: "
+									"<logical(%lu) counter(%u)>",
+									persistedMaxClock.logical,
+									persistedMaxClock.counter)));
+		}
+	}
+
+	LogicalClockShmem->clockInitialized = CLOCKSTATE_INITIALIZED;
+
+	LWLockRelease(&LogicalClockShmem->clockLock);
+}
+
+
+/*
+ * citus_get_node_clock() is an UDF that returns a monotonically increasing
+ * logical clock. Clock guarantees to never go back in value after restarts, and
+ * makes best attempt to keep the value close to unix epoch time in milliseconds.
+ */
+Datum
+citus_get_node_clock(PG_FUNCTION_ARGS)
+{
+	CheckCitusVersion(ERROR);
+
+	ClusterClock *nodeClockValue = (ClusterClock *) palloc(sizeof(ClusterClock));
+
+	GetNextNodeClockValue(nodeClockValue);
+
+	PG_RETURN_POINTER(nodeClockValue);
+}
+
+
+/*
+ * citus_internal_adjust_local_clock_to_remote is an internal UDF to adjust
+ * the local clock to the highest in the cluster.
+ */
+Datum
+citus_internal_adjust_local_clock_to_remote(PG_FUNCTION_ARGS)
+{
+	CheckCitusVersion(ERROR);
+
+	ClusterClock *remoteClock = (ClusterClock *) PG_GETARG_POINTER(0);
+	AdjustLocalClock(remoteClock);
+
+	PG_RETURN_VOID();
+}
+
+
+/*
+ * citus_is_clock_after is an UDF that accepts logical clock timestamps of
+ * two causally related events and returns true if the argument1 happened
+ * before argument2.
+ */
+Datum
+citus_is_clock_after(PG_FUNCTION_ARGS)
+{
+	CheckCitusVersion(ERROR);
+
+	/* Fetch both the arguments */
+	ClusterClock *clock1 = (ClusterClock *) PG_GETARG_POINTER(0);
+	ClusterClock *clock2 = (ClusterClock *) PG_GETARG_POINTER(1);
+
+	ereport(DEBUG1, (errmsg(
+						 "clock1 @ LC:%lu, C:%u, "
+						 "clock2 @ LC:%lu, C:%u",
+						 clock1->logical, clock1->counter,
+						 clock2->logical, clock2->counter)));
+
+	bool result = (cluster_clock_cmp_internal(clock1, clock2) > 0);
+
+	PG_RETURN_BOOL(result);
+}
+
+
+/*
+ * citus_get_transaction_clock() is an UDF that returns a transaction timestamp
+ * logical clock. Clock returned is the maximum of all transaction-nodes and the
+ * all the nodes adjust to the this new clock value.
+ */
+Datum
+citus_get_transaction_clock(PG_FUNCTION_ARGS)
+{
+	CheckCitusVersion(ERROR);
+
+	ClusterClock *clusterClockValue = PrepareAndSetTransactionClock();
+
+	PG_RETURN_POINTER(clusterClockValue);
+}

src/backend/distributed/commands/alter_table.c

@@ -34,9 +34,16 @@
 #include "catalog/pg_am.h"
 #include "catalog/pg_depend.h"
 #include "catalog/pg_rewrite_d.h"
+#include "commands/defrem.h"
+#include "executor/spi.h"
+#include "nodes/pg_list.h"
+#include "utils/builtins.h"
+#include "utils/lsyscache.h"
+#include "utils/syscache.h"
+
 #include "columnar/columnar.h"
 #include "columnar/columnar_tableam.h"
-#include "commands/defrem.h"
+
 #include "distributed/colocation_utils.h"
 #include "distributed/commands.h"
 #include "distributed/commands/utility_hook.h"
@@ -53,17 +60,15 @@
 #include "distributed/multi_executor.h"
 #include "distributed/multi_logical_planner.h"
 #include "distributed/multi_partitioning_utils.h"
+#include "distributed/namespace_utils.h"
 #include "distributed/reference_table_utils.h"
 #include "distributed/relation_access_tracking.h"
-#include "distributed/shared_library_init.h"
+#include "distributed/replication_origin_session_utils.h"
 #include "distributed/shard_utils.h"
+#include "distributed/shared_library_init.h"
+#include "distributed/tenant_schema_metadata.h"
 #include "distributed/worker_protocol.h"
 #include "distributed/worker_transaction.h"
-#include "executor/spi.h"
-#include "nodes/pg_list.h"
-#include "utils/builtins.h"
-#include "utils/lsyscache.h"
-#include "utils/syscache.h"
 
 
 /* Table Conversion Types */
@@ -183,6 +188,7 @@ static TableConversionReturn * AlterDistributedTable(TableConversionParameters *
 static TableConversionReturn * AlterTableSetAccessMethod(
 	TableConversionParameters *params);
 static TableConversionReturn * ConvertTable(TableConversionState *con);
+static TableConversionReturn * ConvertTableInternal(TableConversionState *con);
 static bool SwitchToSequentialAndLocalExecutionIfShardNameTooLong(char *relationName,
 																  char *longestShardName);
 static void DropIndexesNotSupportedByColumnar(Oid relationId,
@@ -194,6 +200,7 @@ static void EnsureTableNotReferencing(Oid relationId, char conversionType);
 static void EnsureTableNotReferenced(Oid relationId, char conversionType);
 static void EnsureTableNotForeign(Oid relationId);
 static void EnsureTableNotPartition(Oid relationId);
+static void ErrorIfColocateWithTenantTable(char *colocateWith);
 static TableConversionState * CreateTableConversion(TableConversionParameters *params);
 static void CreateDistributedTableLike(TableConversionState *con);
 static void CreateCitusTableLike(TableConversionState *con);
@@ -202,12 +209,9 @@ static void ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommand
 static bool HasAnyGeneratedStoredColumns(Oid relationId);
 static List * GetNonGeneratedStoredColumnNameList(Oid relationId);
 static void CheckAlterDistributedTableConversionParameters(TableConversionState *con);
-static char * CreateWorkerChangeSequenceDependencyCommand(char *sequenceSchemaName,
-														  char *sequenceName,
-														  char *sourceSchemaName,
-														  char *sourceName,
-														  char *targetSchemaName,
-														  char *targetName);
+static char * CreateWorkerChangeSequenceDependencyCommand(char *qualifiedSequeceName,
+														  char *qualifiedSourceName,
+														  char *qualifiedTargetName);
 static void ErrorIfMatViewSizeExceedsTheLimit(Oid matViewOid);
 static char * CreateMaterializedViewDDLCommand(Oid matViewOid);
 static char * GetAccessMethodForMatViewIfExists(Oid viewOid);
@@ -215,7 +219,10 @@ static bool WillRecreateForeignKeyToReferenceTable(Oid relationId,
 												   CascadeToColocatedOption cascadeOption);
 static void WarningsForDroppingForeignKeysWithDistributedTables(Oid relationId);
 static void ErrorIfUnsupportedCascadeObjects(Oid relationId);
+static List * WrapTableDDLCommands(List *commandStrings);
 static bool DoesCascadeDropUnsupportedObject(Oid classId, Oid id, HTAB *nodeMap);
+static TableConversionReturn * CopyTableConversionReturnIntoCurrentContext(
+	TableConversionReturn *tableConversionReturn);
 
 PG_FUNCTION_INFO_V1(undistribute_table);
 PG_FUNCTION_INFO_V1(alter_distributed_table);
@@ -242,7 +249,8 @@ undistribute_table(PG_FUNCTION_ARGS)
 
 	TableConversionParameters params = {
 		.relationId = relationId,
-		.cascadeViaForeignKeys = cascadeViaForeignKeys
+		.cascadeViaForeignKeys = cascadeViaForeignKeys,
+		.bypassTenantCheck = false
 	};
 
 	UndistributeTable(&params);
@@ -355,6 +363,124 @@ worker_change_sequence_dependency(PG_FUNCTION_ARGS)
 }
 
 
+/*
+ * DropFKeysAndUndistributeTable drops all foreign keys that relation with
+ * relationId is involved then undistributes it.
+ * Note that as UndistributeTable changes relationId of relation, this
+ * function also returns new relationId of relation.
+ * Also note that callers are responsible for storing & recreating foreign
+ * keys to be dropped if needed.
+ */
+Oid
+DropFKeysAndUndistributeTable(Oid relationId)
+{
+	DropFKeysRelationInvolvedWithTableType(relationId, INCLUDE_ALL_TABLE_TYPES);
+
+	/* store them before calling UndistributeTable as it changes relationId */
+	char *relationName = get_rel_name(relationId);
+	Oid schemaId = get_rel_namespace(relationId);
+
+	/* suppress notices messages not to be too verbose */
+	TableConversionParameters params = {
+		.relationId = relationId,
+		.cascadeViaForeignKeys = false,
+		.suppressNoticeMessages = true
+	};
+	UndistributeTable(&params);
+
+	Oid newRelationId = get_relname_relid(relationName, schemaId);
+
+	/*
+	 * We don't expect this to happen but to be on the safe side let's error
+	 * out here.
+	 */
+	EnsureRelationExists(newRelationId);
+
+	return newRelationId;
+}
+
+
+/*
+ * UndistributeTables undistributes given relations. It first collects all foreign keys
+ * to recreate them after the undistribution. Then, drops the foreign keys and
+ * undistributes the relations. Finally, it recreates foreign keys.
+ */
+void
+UndistributeTables(List *relationIdList)
+{
+	/*
+	 * Collect foreign keys for recreation and then drop fkeys and undistribute
+	 * tables.
+	 */
+	List *originalForeignKeyRecreationCommands = NIL;
+	Oid relationId = InvalidOid;
+	foreach_declared_oid(relationId, relationIdList)
+	{
+		List *fkeyCommandsForRelation =
+			GetFKeyCreationCommandsRelationInvolvedWithTableType(relationId,
+																 INCLUDE_ALL_TABLE_TYPES);
+		originalForeignKeyRecreationCommands = list_concat(
+			originalForeignKeyRecreationCommands, fkeyCommandsForRelation);
+		DropFKeysAndUndistributeTable(relationId);
+	}
+
+	/* We can skip foreign key validations as we are sure about them at start */
+	bool skip_validation = true;
+	ExecuteForeignKeyCreateCommandList(originalForeignKeyRecreationCommands,
+									   skip_validation);
+}
+
+
+/*
+ * EnsureUndistributeTenantTableSafe ensures that it is safe to undistribute a tenant table.
+ */
+void
+EnsureUndistributeTenantTableSafe(Oid relationId, const char *operationName)
+{
+	Oid schemaId = get_rel_namespace(relationId);
+	Assert(IsTenantSchema(schemaId));
+
+	/* We only allow undistribute while altering schema */
+	if (strcmp(operationName, TenantOperationNames[TENANT_SET_SCHEMA]) != 0)
+	{
+		ErrorIfTenantTable(relationId, operationName);
+	}
+
+	char *tableName = get_rel_name(relationId);
+	char *schemaName = get_namespace_name(schemaId);
+
+	/*
+	 * Partition table cannot be undistributed. Otherwise, its parent table would still
+	 * be a tenant table whereas partition table would be a local table.
+	 */
+	if (PartitionTable(relationId))
+	{
+		ereport(ERROR, (errmsg("%s is not allowed for partition table %s in distributed "
+							   "schema %s", operationName, tableName, schemaName),
+						errdetail("partition table should be under the same distributed "
+								  "schema as its parent and be a "
+								  "distributed schema table.")));
+	}
+
+	/*
+	 * When table is referenced by or referencing to a table in the same tenant
+	 * schema, we should disallow undistributing the table since we do not allow
+	 * foreign keys from/to Citus local or Postgres local table to/from distributed
+	 * schema.
+	 */
+	List *fkeyCommandsWithSingleShardTables =
+		GetFKeyCreationCommandsRelationInvolvedWithTableType(
+			relationId, INCLUDE_SINGLE_SHARD_TABLES);
+	if (fkeyCommandsWithSingleShardTables != NIL)
+	{
+		ereport(ERROR, (errmsg("%s is not allowed for table %s in distributed schema %s",
+							   operationName, tableName, schemaName),
+						errdetail("distributed schemas cannot have foreign keys from/to "
+								  "local tables or different schema")));
+	}
+}
+
+
 /*
  * UndistributeTable undistributes the given table. It uses ConvertTable function to
  * create a new local table and move everything to that table.
@@ -375,6 +501,14 @@ UndistributeTable(TableConversionParameters *params)
 							   "because the table is not distributed")));
 	}
 
+	Oid schemaId = get_rel_namespace(params->relationId);
+	if (!params->bypassTenantCheck && IsTenantSchema(schemaId) &&
+		IsCitusTableType(params->relationId, SINGLE_SHARD_DISTRIBUTED))
+	{
+		EnsureUndistributeTenantTableSafe(params->relationId,
+										  TenantOperationNames[TENANT_UNDISTRIBUTE_TABLE]);
+	}
+
 	if (!params->cascadeViaForeignKeys)
 	{
 		EnsureTableNotReferencing(params->relationId, UNDISTRIBUTE_TABLE);
@@ -402,7 +536,11 @@ UndistributeTable(TableConversionParameters *params)
 	params->conversionType = UNDISTRIBUTE_TABLE;
 	params->shardCountIsNull = true;
 	TableConversionState *con = CreateTableConversion(params);
-	return ConvertTable(con);
+
+	SetupReplicationOriginLocalSession();
+	TableConversionReturn *conv = ConvertTable(con);
+	ResetReplicationOriginLocalSession();
+	return conv;
 }
 
 
@@ -426,6 +564,9 @@ AlterDistributedTable(TableConversionParameters *params)
 							   "is not distributed")));
 	}
 
+	ErrorIfTenantTable(params->relationId, TenantOperationNames[TENANT_ALTER_TABLE]);
+	ErrorIfColocateWithTenantTable(params->colocateWith);
+
 	EnsureTableNotForeign(params->relationId);
 	EnsureTableNotPartition(params->relationId);
 	EnsureHashDistributedTable(params->relationId);
@@ -441,6 +582,7 @@ AlterDistributedTable(TableConversionParameters *params)
 		ereport(DEBUG1, (errmsg("setting multi shard modify mode to sequential")));
 		SetLocalMultiShardModifyModeToSequential();
 	}
+
 	return ConvertTable(con);
 }
 
@@ -467,8 +609,11 @@ AlterTableSetAccessMethod(TableConversionParameters *params)
 	EnsureTableNotReferencing(params->relationId, ALTER_TABLE_SET_ACCESS_METHOD);
 	EnsureTableNotReferenced(params->relationId, ALTER_TABLE_SET_ACCESS_METHOD);
 	EnsureTableNotForeign(params->relationId);
-	if (IsCitusTableType(params->relationId, DISTRIBUTED_TABLE))
+
+	if (!IsCitusTableType(params->relationId, SINGLE_SHARD_DISTRIBUTED) &&
+		IsCitusTableType(params->relationId, DISTRIBUTED_TABLE))
 	{
+		/* we do not support non-hash distributed tables, except single shard tables */
 		EnsureHashDistributedTable(params->relationId);
 	}
 
@@ -477,6 +622,11 @@ AlterTableSetAccessMethod(TableConversionParameters *params)
 		ereport(ERROR, (errmsg("you cannot alter access method of a partitioned table")));
 	}
 
+	if (get_rel_relkind(params->relationId) == RELKIND_VIEW)
+	{
+		ereport(ERROR, (errmsg("you cannot alter access method of a view")));
+	}
+
 	if (PartitionTable(params->relationId) &&
 		IsCitusTableType(params->relationId, DISTRIBUTED_TABLE))
 	{
@@ -506,9 +656,9 @@ AlterTableSetAccessMethod(TableConversionParameters *params)
 
 
 /*
- * ConvertTable is used for converting a table into a new table with different properties.
- * The conversion is done by creating a new table, moving everything to the new table and
- * dropping the old one. So the oid of the table is not preserved.
+ * ConvertTableInternal is used for converting a table into a new table with different
+ * properties. The conversion is done by creating a new table, moving everything to the
+ * new table and dropping the old one. So the oid of the table is not preserved.
  *
  * The new table will have the same name, columns and rows. It will also have partitions,
  * views, sequences of the old table. Finally it will have everything created by
@@ -527,7 +677,7 @@ AlterTableSetAccessMethod(TableConversionParameters *params)
  * in case you add a new way to return from this function.
  */
 TableConversionReturn *
-ConvertTable(TableConversionState *con)
+ConvertTableInternal(TableConversionState *con)
 {
 	InTableTypeConversionFunctionCall = true;
 
@@ -565,8 +715,11 @@ ConvertTable(TableConversionState *con)
 	char *newAccessMethod = con->accessMethod ? con->accessMethod :
 							con->originalAccessMethod;
 	IncludeSequenceDefaults includeSequenceDefaults = NEXTVAL_SEQUENCE_DEFAULTS;
+	IncludeIdentities includeIdentity = INCLUDE_IDENTITY;
+
 	List *preLoadCommands = GetPreLoadTableCreationCommands(con->relationId,
 															includeSequenceDefaults,
+															includeIdentity,
 															newAccessMethod);
 
 	if (con->accessMethod && strcmp(con->accessMethod, "columnar") == 0)
@@ -587,9 +740,18 @@ ConvertTable(TableConversionState *con)
 	List *justBeforeDropCommands = NIL;
 	List *attachPartitionCommands = NIL;
 
-	postLoadCommands =
-		list_concat(postLoadCommands,
-					GetViewCreationTableDDLCommandsOfTable(con->relationId));
+	List *createViewCommands = GetViewCreationCommandsOfTable(con->relationId);
+
+	postLoadCommands = list_concat(postLoadCommands,
+								   WrapTableDDLCommands(createViewCommands));
+
+	/* need to add back to publications after dropping the original table */
+	bool isAdd = true;
+	List *alterPublicationCommands =
+		GetAlterPublicationDDLCommandsForTable(con->relationId, isAdd);
+
+	postLoadCommands = list_concat(postLoadCommands,
+								   WrapTableDDLCommands(alterPublicationCommands));
 
 	List *foreignKeyCommands = NIL;
 	if (con->conversionType == ALTER_DISTRIBUTED_TABLE)
@@ -626,19 +788,21 @@ ConvertTable(TableConversionState *con)
 		justBeforeDropCommands = lappend(justBeforeDropCommands, detachFromParentCommand);
 	}
 
+	char *qualifiedRelationName = quote_qualified_identifier(con->schemaName,
+															 con->relationName);
+
 	if (PartitionedTable(con->relationId))
 	{
 		if (!con->suppressNoticeMessages)
 		{
 			ereport(NOTICE, (errmsg("converting the partitions of %s",
-									quote_qualified_identifier(con->schemaName,
-															   con->relationName))));
+									qualifiedRelationName)));
 		}
 
 		List *partitionList = PartitionList(con->relationId);
 
 		Oid partitionRelationId = InvalidOid;
-		foreach_oid(partitionRelationId, partitionList)
+		foreach_declared_oid(partitionRelationId, partitionList)
 		{
 			char *tableQualifiedName = generate_qualified_relation_name(
 				partitionRelationId);
@@ -705,13 +869,11 @@ ConvertTable(TableConversionState *con)
 
 	if (!con->suppressNoticeMessages)
 	{
-		ereport(NOTICE, (errmsg("creating a new table for %s",
-								quote_qualified_identifier(con->schemaName,
-														   con->relationName))));
+		ereport(NOTICE, (errmsg("creating a new table for %s", qualifiedRelationName)));
 	}
 
 	TableDDLCommand *tableCreationCommand = NULL;
-	foreach_ptr(tableCreationCommand, preLoadCommands)
+	foreach_declared_ptr(tableCreationCommand, preLoadCommands)
 	{
 		Assert(CitusIsA(tableCreationCommand, TableDDLCommand));
 
@@ -785,16 +947,28 @@ ConvertTable(TableConversionState *con)
 				 con->suppressNoticeMessages);
 
 	TableDDLCommand *tableConstructionCommand = NULL;
-	foreach_ptr(tableConstructionCommand, postLoadCommands)
+	foreach_declared_ptr(tableConstructionCommand, postLoadCommands)
 	{
 		Assert(CitusIsA(tableConstructionCommand, TableDDLCommand));
 		char *tableConstructionSQL = GetTableDDLCommand(tableConstructionCommand);
 		ExecuteQueryViaSPI(tableConstructionSQL, SPI_OK_UTILITY);
 	}
 
+	/*
+	 * when there are many partitions, each call to ProcessUtilityParseTree
+	 * accumulates used memory. Free context after each call.
+	 */
+	MemoryContext citusPerPartitionContext =
+		AllocSetContextCreate(CurrentMemoryContext,
+							  "citus_per_partition_context",
+							  ALLOCSET_DEFAULT_SIZES);
+	MemoryContext oldContext = MemoryContextSwitchTo(citusPerPartitionContext);
+
 	char *attachPartitionCommand = NULL;
-	foreach_ptr(attachPartitionCommand, attachPartitionCommands)
+	foreach_declared_ptr(attachPartitionCommand, attachPartitionCommands)
 	{
+		MemoryContextReset(citusPerPartitionContext);
+
 		Node *parseTree = ParseTreeNode(attachPartitionCommand);
 
 		ProcessUtilityParseTree(parseTree, attachPartitionCommand,
@@ -802,6 +976,9 @@ ConvertTable(TableConversionState *con)
 								NULL, None_Receiver, NULL);
 	}
 
+	MemoryContextSwitchTo(oldContext);
+	MemoryContextDelete(citusPerPartitionContext);
+
 	if (isPartitionTable)
 	{
 		ExecuteQueryViaSPI(attachToParentCommand, SPI_OK_UTILITY);
@@ -813,14 +990,12 @@ ConvertTable(TableConversionState *con)
 
 		/* For now we only support cascade to colocation for alter_distributed_table UDF */
 		Assert(con->conversionType == ALTER_DISTRIBUTED_TABLE);
-		foreach_oid(colocatedTableId, con->colocatedTableList)
+		foreach_declared_oid(colocatedTableId, con->colocatedTableList)
 		{
 			if (colocatedTableId == con->relationId)
 			{
 				continue;
 			}
-			char *qualifiedRelationName = quote_qualified_identifier(con->schemaName,
-																	 con->relationName);
 
 			TableConversionParameters cascadeParam = {
 				.relationId = colocatedTableId,
@@ -843,7 +1018,7 @@ ConvertTable(TableConversionState *con)
 		if (con->cascadeToColocated != CASCADE_TO_COLOCATED_NO_ALREADY_CASCADED)
 		{
 			char *foreignKeyCommand = NULL;
-			foreach_ptr(foreignKeyCommand, foreignKeyCommands)
+			foreach_declared_ptr(foreignKeyCommand, foreignKeyCommands)
 			{
 				ExecuteQueryViaSPI(foreignKeyCommand, SPI_OK_UTILITY);
 			}
@@ -861,10 +1036,77 @@ ConvertTable(TableConversionState *con)
 	SetLocalEnableLocalReferenceForeignKeys(oldEnableLocalReferenceForeignKeys);
 
 	InTableTypeConversionFunctionCall = false;
+
 	return ret;
 }
 
 
+/*
+ * CopyTableConversionReturnIntoCurrentContext copies given tableConversionReturn
+ * into CurrentMemoryContext.
+ */
+static TableConversionReturn *
+CopyTableConversionReturnIntoCurrentContext(TableConversionReturn *tableConversionReturn)
+{
+	TableConversionReturn *tableConversionReturnCopy = NULL;
+	if (tableConversionReturn)
+	{
+		tableConversionReturnCopy = palloc0(sizeof(TableConversionReturn));
+		List *copyForeignKeyCommands = NIL;
+		char *foreignKeyCommand = NULL;
+		foreach_declared_ptr(foreignKeyCommand, tableConversionReturn->foreignKeyCommands)
+		{
+			char *copyForeignKeyCommand = MemoryContextStrdup(CurrentMemoryContext,
+															  foreignKeyCommand);
+			copyForeignKeyCommands = lappend(copyForeignKeyCommands,
+											 copyForeignKeyCommand);
+		}
+		tableConversionReturnCopy->foreignKeyCommands = copyForeignKeyCommands;
+	}
+
+	return tableConversionReturnCopy;
+}
+
+
+/*
+ * ConvertTable is a wrapper for ConvertTableInternal to persist only
+ * TableConversionReturn and delete all other allocations.
+ */
+static TableConversionReturn *
+ConvertTable(TableConversionState *con)
+{
+	/*
+	 * We do not allow alter_distributed_table and undistribute_table operations
+	 * for tables with identity columns. This is because we do not have a proper way
+	 * of keeping sequence states consistent across the cluster.
+	 */
+	ErrorIfTableHasIdentityColumn(con->relationId);
+
+	/*
+	 * when there are many partitions or colocated tables, memory usage is
+	 * accumulated. Free context for each call to ConvertTable.
+	 */
+	MemoryContext convertTableContext =
+		AllocSetContextCreate(CurrentMemoryContext,
+							  "citus_convert_table_context",
+							  ALLOCSET_DEFAULT_SIZES);
+	MemoryContext oldContext = MemoryContextSwitchTo(convertTableContext);
+
+	TableConversionReturn *tableConversionReturn = ConvertTableInternal(con);
+
+	MemoryContextSwitchTo(oldContext);
+
+	/* persist TableConversionReturn in oldContext */
+	TableConversionReturn *tableConversionReturnCopy =
+		CopyTableConversionReturnIntoCurrentContext(tableConversionReturn);
+
+	/* delete convertTableContext */
+	MemoryContextDelete(convertTableContext);
+
+	return tableConversionReturnCopy;
+}
+
+
 /*
  * DropIndexesNotSupportedByColumnar is a helper function used during accces
  * method conversion to drop the indexes that are not supported by columnarAM.
@@ -887,7 +1129,7 @@ DropIndexesNotSupportedByColumnar(Oid relationId, bool suppressNoticeMessages)
 	RelationClose(columnarRelation);
 
 	Oid indexId = InvalidOid;
-	foreach_oid(indexId, indexIdList)
+	foreach_declared_oid(indexId, indexIdList)
 	{
 		char *indexAmName = GetIndexAccessMethodName(indexId);
 		if (extern_ColumnarSupportsIndexAM(indexAmName))
@@ -1068,6 +1310,25 @@ EnsureTableNotPartition(Oid relationId)
 }
 
 
+/*
+ * ErrorIfColocateWithTenantTable errors out if given colocateWith text refers to
+ * a tenant table.
+ */
+void
+ErrorIfColocateWithTenantTable(char *colocateWith)
+{
+	if (colocateWith != NULL &&
+		!IsColocateWithDefault(colocateWith) &&
+		!IsColocateWithNone(colocateWith))
+	{
+		text *colocateWithTableNameText = cstring_to_text(colocateWith);
+		Oid colocateWithTableId = ResolveRelationId(colocateWithTableNameText, false);
+		ErrorIfTenantTable(colocateWithTableId,
+						   TenantOperationNames[TENANT_COLOCATE_WITH]);
+	}
+}
+
+
 TableConversionState *
 CreateTableConversion(TableConversionParameters *params)
 {
@@ -1091,19 +1352,7 @@ CreateTableConversion(TableConversionParameters *params)
 							   "because no such table exists")));
 	}
 
-	TupleDesc relationDesc = RelationGetDescr(relation);
-	if (RelationUsesIdentityColumns(relationDesc))
-	{
-		/*
-		 * pg_get_tableschemadef_string doesn't know how to deparse identity
-		 * columns so we cannot reflect those columns when creating table
-		 * from scratch. For this reason, error out here.
-		 */
-		ereport(ERROR, (errmsg("cannot complete command because relation "
-							   "%s has identity column",
-							   generate_qualified_relation_name(con->relationId)),
-						errhint("Drop the identity columns and re-try the command")));
-	}
+
 	relation_close(relation, NoLock);
 	con->distributionKey =
 		BuildDistributionKeyFromColumnName(con->relationId, con->distributionColumn,
@@ -1140,7 +1389,7 @@ CreateTableConversion(TableConversionParameters *params)
 		 * since they will be handled separately.
 		 */
 		Oid colocatedTableId = InvalidOid;
-		foreach_oid(colocatedTableId, colocatedTableList)
+		foreach_declared_oid(colocatedTableId, colocatedTableList)
 		{
 			if (PartitionTable(colocatedTableId))
 			{
@@ -1223,8 +1472,15 @@ CreateDistributedTableLike(TableConversionState *con)
 		newShardCount = con->shardCount;
 	}
 
+	/*
+	 * To get the correct column name, we use the original relation id, not the
+	 * new relation id. The reason is that the cached attributes of the original
+	 * and newly created tables are not the same if the original table has
+	 * dropped columns (dropped columns are still present in the attribute cache)
+	 * Detailed example in https://github.com/citusdata/citus/pull/6387
+	 */
 	char *distributionColumnName =
-		ColumnToColumnName(con->newRelationId, (Node *) newDistributionKey);
+		ColumnToColumnName(con->relationId, (Node *) newDistributionKey);
 
 	Oid originalRelationId = con->relationId;
 	if (con->originalDistributionKey != NULL && PartitionTable(originalRelationId))
@@ -1248,7 +1504,7 @@ CreateDistributedTableLike(TableConversionState *con)
 	char partitionMethod = PartitionMethod(con->relationId);
 
 	CreateDistributedTable(con->newRelationId, distributionColumnName, partitionMethod,
-						   newShardCount, true, newColocateWith, false);
+						   newShardCount, true, newColocateWith);
 }
 
 
@@ -1261,12 +1517,23 @@ CreateCitusTableLike(TableConversionState *con)
 {
 	if (IsCitusTableType(con->relationId, DISTRIBUTED_TABLE))
 	{
-		CreateDistributedTableLike(con);
+		if (IsCitusTableType(con->relationId, SINGLE_SHARD_DISTRIBUTED))
+		{
+			ColocationParam colocationParam = {
+				.colocationParamType = COLOCATE_WITH_TABLE_LIKE_OPT,
+				.colocateWithTableName = quote_qualified_identifier(con->schemaName,
+																	con->relationName)
+			};
+			CreateSingleShardTable(con->newRelationId, colocationParam);
+		}
+		else
+		{
+			CreateDistributedTableLike(con);
+		}
 	}
 	else if (IsCitusTableType(con->relationId, REFERENCE_TABLE))
 	{
-		CreateDistributedTable(con->newRelationId, NULL, DISTRIBUTE_BY_NONE, 0, false,
-							   NULL, false);
+		CreateReferenceTable(con->newRelationId);
 	}
 	else if (IsCitusTableType(con->relationId, CITUS_LOCAL_TABLE))
 	{
@@ -1338,7 +1605,7 @@ DoesCascadeDropUnsupportedObject(Oid classId, Oid objectId, HTAB *nodeMap)
 																	 targetObjectId);
 
 	HeapTuple depTup = NULL;
-	foreach_ptr(depTup, dependencyTupleList)
+	foreach_declared_ptr(depTup, dependencyTupleList)
 	{
 		Form_pg_depend pg_depend = (Form_pg_depend) GETSTRUCT(depTup);
 
@@ -1378,7 +1645,7 @@ GetViewCreationCommandsOfTable(Oid relationId)
 	List *commands = NIL;
 
 	Oid viewOid = InvalidOid;
-	foreach_oid(viewOid, views)
+	foreach_declared_oid(viewOid, views)
 	{
 		StringInfo query = makeStringInfo();
 
@@ -1407,17 +1674,16 @@ GetViewCreationCommandsOfTable(Oid relationId)
 
 
 /*
- * GetViewCreationTableDDLCommandsOfTable is the same as GetViewCreationCommandsOfTable,
- * but the returned list includes objects of TableDDLCommand's, not strings.
+ * WrapTableDDLCommands takes a list of command strings and wraps them
+ * in TableDDLCommand structs.
  */
-List *
-GetViewCreationTableDDLCommandsOfTable(Oid relationId)
+static List *
+WrapTableDDLCommands(List *commandStrings)
 {
-	List *commands = GetViewCreationCommandsOfTable(relationId);
 	List *tableDDLCommands = NIL;
 
 	char *command = NULL;
-	foreach_ptr(command, commands)
+	foreach_declared_ptr(command, commandStrings)
 	{
 		tableDDLCommands = lappend(tableDDLCommands, makeTableDDLCommandString(command));
 	}
@@ -1479,9 +1745,7 @@ CreateMaterializedViewDDLCommand(Oid matViewOid)
 {
 	StringInfo query = makeStringInfo();
 
-	char *viewName = get_rel_name(matViewOid);
-	char *schemaName = get_namespace_name(get_rel_namespace(matViewOid));
-	char *qualifiedViewName = quote_qualified_identifier(schemaName, viewName);
+	char *qualifiedViewName = generate_qualified_relation_name(matViewOid);
 
 	/* here we need to get the access method of the view to recreate it */
 	char *accessMethodName = GetAccessMethodForMatViewIfExists(matViewOid);
@@ -1497,10 +1761,7 @@ CreateMaterializedViewDDLCommand(Oid matViewOid)
 	 * Set search_path to NIL so that all objects outside of pg_catalog will be
 	 * schema-prefixed.
 	 */
-	OverrideSearchPath *overridePath = GetOverrideSearchPath(CurrentMemoryContext);
-	overridePath->schemas = NIL;
-	overridePath->addCatalog = true;
-	PushOverrideSearchPath(overridePath);
+	int saveNestLevel = PushEmptySearchPath();
 
 	/*
 	 * Push the transaction snapshot to be able to get vief definition with pg_get_viewdef
@@ -1512,7 +1773,7 @@ CreateMaterializedViewDDLCommand(Oid matViewOid)
 	char *viewDefinition = TextDatumGetCString(viewDefinitionDatum);
 
 	PopActiveSnapshot();
-	PopOverrideSearchPath();
+	PopEmptySearchPath(saveNestLevel);
 
 	appendStringInfo(query, "AS %s", viewDefinition);
 
@@ -1533,9 +1794,8 @@ ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommands,
 			 bool suppressNoticeMessages)
 {
 	char *sourceName = get_rel_name(sourceId);
-	char *targetName = get_rel_name(targetId);
-	Oid schemaId = get_rel_namespace(sourceId);
-	char *schemaName = get_namespace_name(schemaId);
+	char *qualifiedSourceName = generate_qualified_relation_name(sourceId);
+	char *qualifiedTargetName = generate_qualified_relation_name(targetId);
 
 	StringInfo query = makeStringInfo();
 
@@ -1543,8 +1803,7 @@ ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommands,
 	{
 		if (!suppressNoticeMessages)
 		{
-			ereport(NOTICE, (errmsg("moving the data of %s",
-									quote_qualified_identifier(schemaName, sourceName))));
+			ereport(NOTICE, (errmsg("moving the data of %s", qualifiedSourceName)));
 		}
 
 		if (!HasAnyGeneratedStoredColumns(sourceId))
@@ -1554,8 +1813,7 @@ ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommands,
 			 * "INSERT INTO .. SELECT *"".
 			 */
 			appendStringInfo(query, "INSERT INTO %s SELECT * FROM %s",
-							 quote_qualified_identifier(schemaName, targetName),
-							 quote_qualified_identifier(schemaName, sourceName));
+							 qualifiedTargetName, qualifiedSourceName);
 		}
 		else
 		{
@@ -1568,18 +1826,21 @@ ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommands,
 			 */
 			List *nonStoredColumnNameList = GetNonGeneratedStoredColumnNameList(sourceId);
 			char *insertColumnString = StringJoin(nonStoredColumnNameList, ',');
-			appendStringInfo(query, "INSERT INTO %s (%s) SELECT %s FROM %s",
-							 quote_qualified_identifier(schemaName, targetName),
-							 insertColumnString, insertColumnString,
-							 quote_qualified_identifier(schemaName, sourceName));
+			appendStringInfo(query,
+							 "INSERT INTO %s (%s) OVERRIDING SYSTEM VALUE SELECT %s FROM %s",
+							 qualifiedTargetName, insertColumnString,
+							 insertColumnString, qualifiedSourceName);
 		}
 
 		ExecuteQueryViaSPI(query->data, SPI_OK_INSERT);
 	}
 
-	List *ownedSequences = getOwnedSequences(sourceId);
+	/*
+	 * Modify regular sequence dependencies (sequences marked as DEPENDENCY_AUTO)
+	 */
+	List *ownedSequences = getOwnedSequences_internal(sourceId, 0, DEPENDENCY_AUTO);
 	Oid sequenceOid = InvalidOid;
-	foreach_oid(sequenceOid, ownedSequences)
+	foreach_declared_oid(sequenceOid, ownedSequences)
 	{
 		changeDependencyFor(RelationRelationId, sequenceOid,
 							RelationRelationId, sourceId, targetId);
@@ -1592,14 +1853,11 @@ ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommands,
 		 */
 		if (ShouldSyncTableMetadata(targetId))
 		{
-			Oid sequenceSchemaOid = get_rel_namespace(sequenceOid);
-			char *sequenceSchemaName = get_namespace_name(sequenceSchemaOid);
-			char *sequenceName = get_rel_name(sequenceOid);
+			char *qualifiedSequenceName = generate_qualified_relation_name(sequenceOid);
 			char *workerChangeSequenceDependencyCommand =
-				CreateWorkerChangeSequenceDependencyCommand(sequenceSchemaName,
-															sequenceName,
-															schemaName, sourceName,
-															schemaName, targetName);
+				CreateWorkerChangeSequenceDependencyCommand(qualifiedSequenceName,
+															qualifiedSourceName,
+															qualifiedTargetName);
 			SendCommandToWorkersWithMetadata(workerChangeSequenceDependencyCommand);
 		}
 		else if (ShouldSyncTableMetadata(sourceId))
@@ -1609,43 +1867,36 @@ ReplaceTable(Oid sourceId, Oid targetId, List *justBeforeDropCommands,
 			 * so we should prevent dropping the sequence on the table. Otherwise, we'd
 			 * lose track of the previous changes in the sequence.
 			 */
-			StringInfo command = makeStringInfo();
-
-			appendStringInfo(command,
-							 "SELECT pg_catalog.worker_drop_sequence_dependency('%s');",
-							 quote_qualified_identifier(schemaName, sourceName));
-
-			SendCommandToWorkersWithMetadata(command->data);
+			char *command = WorkerDropSequenceDependencyCommand(sourceId);
+			SendCommandToWorkersWithMetadata(command);
 		}
 	}
 
 	char *justBeforeDropCommand = NULL;
-	foreach_ptr(justBeforeDropCommand, justBeforeDropCommands)
+	foreach_declared_ptr(justBeforeDropCommand, justBeforeDropCommands)
 	{
 		ExecuteQueryViaSPI(justBeforeDropCommand, SPI_OK_UTILITY);
 	}
 
 	if (!suppressNoticeMessages)
 	{
-		ereport(NOTICE, (errmsg("dropping the old %s",
-								quote_qualified_identifier(schemaName, sourceName))));
+		ereport(NOTICE, (errmsg("dropping the old %s", qualifiedSourceName)));
 	}
 
 	resetStringInfo(query);
 	appendStringInfo(query, "DROP %sTABLE %s CASCADE",
 					 IsForeignTable(sourceId) ? "FOREIGN " : "",
-					 quote_qualified_identifier(schemaName, sourceName));
+					 qualifiedSourceName);
 	ExecuteQueryViaSPI(query->data, SPI_OK_UTILITY);
 
 	if (!suppressNoticeMessages)
 	{
-		ereport(NOTICE, (errmsg("renaming the new table to %s",
-								quote_qualified_identifier(schemaName, sourceName))));
+		ereport(NOTICE, (errmsg("renaming the new table to %s", qualifiedSourceName)));
 	}
 
 	resetStringInfo(query);
 	appendStringInfo(query, "ALTER TABLE %s RENAME TO %s",
-					 quote_qualified_identifier(schemaName, targetName),
+					 qualifiedTargetName,
 					 quote_identifier(sourceName));
 	ExecuteQueryViaSPI(query->data, SPI_OK_UTILITY);
 }
@@ -1736,7 +1987,7 @@ CheckAlterDistributedTableConversionParameters(TableConversionState *con)
 		Oid colocatedTableOid = InvalidOid;
 		text *colocateWithText = cstring_to_text(con->colocateWith);
 		Oid colocateWithTableOid = ResolveRelationId(colocateWithText, false);
-		foreach_oid(colocatedTableOid, con->colocatedTableList)
+		foreach_declared_oid(colocatedTableOid, con->colocatedTableList)
 		{
 			if (colocateWithTableOid == colocatedTableOid)
 			{
@@ -1755,6 +2006,12 @@ CheckAlterDistributedTableConversionParameters(TableConversionState *con)
 								   "it is not a distributed table",
 								   con->colocateWith)));
 		}
+		else if (IsCitusTableType(colocateWithTableOid, SINGLE_SHARD_DISTRIBUTED))
+		{
+			ereport(ERROR, (errmsg("cannot colocate with %s because "
+								   "it is a single shard distributed table",
+								   con->colocateWith)));
+		}
 	}
 
 	/* shard_count:=0 is not allowed */
@@ -1899,15 +2156,16 @@ CheckAlterDistributedTableConversionParameters(TableConversionState *con)
  * worker_change_sequence_dependency query with the parameters.
  */
 static char *
-CreateWorkerChangeSequenceDependencyCommand(char *sequenceSchemaName, char *sequenceName,
-											char *sourceSchemaName, char *sourceName,
-											char *targetSchemaName, char *targetName)
+CreateWorkerChangeSequenceDependencyCommand(char *qualifiedSequeceName,
+											char *qualifiedSourceName,
+											char *qualifiedTargetName)
 {
 	StringInfo query = makeStringInfo();
-	appendStringInfo(query, "SELECT worker_change_sequence_dependency('%s', '%s', '%s')",
-					 quote_qualified_identifier(sequenceSchemaName, sequenceName),
-					 quote_qualified_identifier(sourceSchemaName, sourceName),
-					 quote_qualified_identifier(targetSchemaName, targetName));
+	appendStringInfo(query, "SELECT worker_change_sequence_dependency(%s, %s, %s)",
+					 quote_literal_cstr(qualifiedSequeceName),
+					 quote_literal_cstr(qualifiedSourceName),
+					 quote_literal_cstr(qualifiedTargetName));
+
 	return query->data;
 }
 
@@ -1956,7 +2214,7 @@ WillRecreateForeignKeyToReferenceTable(Oid relationId,
 	{
 		List *colocatedTableList = ColocatedTableList(relationId);
 		Oid colocatedTableOid = InvalidOid;
-		foreach_oid(colocatedTableOid, colocatedTableList)
+		foreach_declared_oid(colocatedTableOid, colocatedTableList)
 		{
 			if (HasForeignKeyToReferenceTable(colocatedTableOid))
 			{
@@ -1984,7 +2242,7 @@ WarningsForDroppingForeignKeysWithDistributedTables(Oid relationId)
 	List *foreignKeys = list_concat(referencingForeingKeys, referencedForeignKeys);
 
 	Oid foreignKeyOid = InvalidOid;
-	foreach_oid(foreignKeyOid, foreignKeys)
+	foreach_declared_oid(foreignKeyOid, foreignKeys)
 	{
 		ereport(WARNING, (errmsg("foreign key %s will be dropped",
 								 get_constraint_name(foreignKeyOid))));