Test isolation_update_node fails on some systems with the following error:
```
-s2: WARNING: connection to the remote node non-existent:57637 failed with the following error: could not translate host name "non-existent" to address: Name or service not known
+s2: WARNING: connection to the remote node non-existent:57637 failed with the following error: could not translate host name "non-existent" to address: Temporary failure in name resolution
```
This slightly modifies an already existing [normalization
rule](739c6d26df/src/test/regress/bin/normalize.sed (L217-L218))
to fix it.
Co-authored-by: Karina Litskevich <litskevichkarina@gmail.com>
Adding upgrade_basic_before_non_mixed.sql file because while
upgrade_basic_after_non_mixed exist, its before variation didn't exist
as we don't have any "before" steps. However, run_test.py assumes that
all "after" files do have a "before" variation as well. So this PR adds
an empty upgrade_basic_before_non_mixed.sql file.
Also, given that we don't have such a version called as 12.1devel
anymore, change it to 12.1.1.
And finally, let CI skip testing flakyness for upgrade tests both
because it's quite hard to get flaky-test-detection job working for
upgrade tests and also because in the end it is not much useful to test
upgrade tests against flakyness.
Running a query from a Citus non-main database that inserts to
pg_dist_object requires a new connection to the main database itself.
This PR adds that connection to the main database.
---------
Co-authored-by: Jelte Fennema-Nio <github-tech@jeltef.nl>
When there are multiple localhost entries in /etc/hosts like following
/etc/hosts:
```
127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
127.0.0.1 localhost
```
multi_cluster_management check will failed:
```
@@ -857,20 +857,21 @@
ERROR: group 14 already has a primary node
-- check that you can add secondaries and unavailable nodes to a group
SELECT groupid AS worker_2_group FROM pg_dist_node WHERE nodeport = :worker_2_port \gset
SELECT 1 FROM master_add_node('localhost', 9998, groupid => :worker_1_group, noderole => 'secondary');
?column?
----------
1
(1 row)
SELECT 1 FROM master_add_node('localhost', 9997, groupid => :worker_1_group, noderole => 'unavailable');
+WARNING: could not establish connection after 5000 ms
?column?
----------
1
(1 row)
```
This actually isn't just a problem in test environments, but could occur
as well during actual usage when a hostname in pg_dist_node
resolves to multiple IPs and one of those IPs is unreachable.
Postgres will then automatically continue with the next IP, but
Citus should listen for events on the new socket. Not on the
old one.
Co-authored-by: chuhx43211 <chuhx43211@hundsun.com>
LoadShardList is called twice, which is not neccessary, and there is no
need to sort the shard placement list since we only want to know the list
length.
DESCRIPTION: Adds support for issuing `CREATE`/`DROP` DATABASE commands
from worker nodes
With this commit, we allow issuing CREATE / DROP DATABASE commands from
worker nodes too.
As in #7278, this is not allowed when the coordinator is not added to
metadata because we don't ever sync metadata changes to coordinator
when adding coordinator to the metadata via
`SELECT citus_set_coordinator_host('<hostname>')`, or equivalently, via
`SELECT citus_add_node(<coordinator_node_name>, <coordinator_node_port>, 0)`.
We serialize database management commands by acquiring a Citus specific
advisory lock on the first primary worker node if there are any workers in the
cluster. As opposed to what we've done in https://github.com/citusdata/citus/pull/7278
for role management commands, we try to avoid from running into distributed deadlocks
as much as possible. This is because, while distributed deadlocks that can happen around
role management commands can be detected by Citus, this is not the case for database
management commands because most of them cannot be run inside in a transaction block.
In that case, Citus cannot even detect the distributed deadlock because the command is not
part of a distributed transaction at all, then the command execution might not return the
control back to the user for an indefinite amount of time.
This fixes#7230.
First of all, using HeapTupleHeaderGetDatumLength(heapTuple) is
definetly wrong, it gives a number that's 4 times less than the correct
tuple size (heapTuple.t_len). See
https://github.com/postgres/postgres/blob/REL_16_0/src/include/access/htup_details.h#L455-L456https://github.com/postgres/postgres/blob/REL_16_0/src/include/varatt.h#L279https://github.com/postgres/postgres/blob/REL_16_0/src/include/varatt.h#L225-L226
When I fixed it, the limit_intermediate_size test failed, so I tried to
understand what's going on there. In original commit fd546cf these
queries were supposed to fail. Then in b3af63c three of the queries that
were supposed to fail suddenly worked and tests were changed to pass
without understanding why the output had changed or how to keep test
testing what it had to test. Even comments saying that these queries
should fail were left untouched. Commit message gives no clue about why
exactly test has changed:
> It seems that when we use adaptive executor instead of task tracker,
we
> exceed the intermediate result size less in the test. Therefore
updated
> the tests accordingly.
Then 3fda2c3 also blindly raised the limit for one of the queries to
keep it working:
3fda2c3254 (diff-a9b7b617f9dfd345318cb8987d5897143ca1b723c87b81049bbadd94dcc86570R19)
When in fe3caf3 that HeapTupleHeaderGetDatumLength(heapTuple) call was
finally added, one of those test queries became failing again.
The other two of them now also failing after the fix. I don't understand
how exactly the calculation of "intermediate result size" that is
limited by citus.max_intermediate_result_size had changed through
b3af63c and fe3caf3, but these numbers are now closer to what
they originally were when this limitation was added in
fd546cf. So these queries should fail, like in the original
version of the limit_intermediate_size test.
Co-authored-by: Karina Litskevich <litskevichkarina@gmail.com>
foreign_key_to_reference_shard_rebalance failed because partition of
2024 year does not exist, fixed by add default partition.
Replaces https://github.com/citusdata/citus/pull/7396 by adding a rule
that allows properly testing foreign_key_to_reference_shard_rebalance
via run_test.py.
Closes#7396
Co-authored-by: chuhx <148182736+cstarc1@users.noreply.github.com>
DESCRIPTION: Adds REASSIGN OWNED BY propagation
This pull request introduces the propagation of the "Reassign owned by"
statement. It accommodates both local and distributed roles for both the
old and new assignments. However, when the old role is a local role, it
undergoes filtering and is not propagated. On the other hand, if the new
role is a local role, the process involves first creating the role on
worker nodes before propagating the "Reassign owned" statement.
DESCRIPTION: Adds database connection limit, rename and set tablespace
propagation
In this PR, below statement propagations are added
alter database <database_name> with allow_connections = <boolean_value>;
alter database <database_name> rename to <database_name2>;
alter database <database_name> set TABLESPACE <table_space_name>
---------
Co-authored-by: Jelte Fennema-Nio <github-tech@jeltef.nl>
Co-authored-by: Jelte Fennema-Nio <jelte.fennema@microsoft.com>
Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>
DESCRIPTION: Adds support for 2PC from non-Citus main databases
This PR only adds support for `CREATE USER` queries, other queries need
to be added. But it should be simple because this PR creates the
underlying structure.
Citus main database is the database where the Citus extension is
created. A non-main database is all the other databases that are in the
same node with a Citus main database.
When a `CREATE USER` query is run on a non-main database we:
1. Run `start_management_transaction` on the main database. This
function saves the outer transaction's xid (the non-main database
query's transaction id) and marks the current query as main db command.
2. Run `execute_command_on_remote_nodes_as_user("CREATE USER
<username>", <username to run the command>)` on the main database. This
function creates the users in the rest of the cluster by running the
query on the other nodes. The user on the current node is created by the
query on the outer, non-main db, query to make sure consequent commands
in the same transaction can see this user.
3. Run `mark_object_distributed` on the main database. This function
adds the user to `pg_dist_object` in all of the nodes, including the
current one.
This PR also implements transaction recovery for the queries from
non-main databases.
Allowing GRANT ADMIN to now also be INHERIT or SET in support of psql16
GRANT role_name [, ...] TO role_specification [, ...] [ WITH { ADMIN |
INHERIT | SET } { OPTION | TRUE | FALSE } ] [ GRANTED BY
role_specification ]
Fixes: #7148
Related: #7138
See review changes from https://github.com/citusdata/citus/pull/7164
This change adds a script to programatically group all includes in a
specific order. The script was used as a one time invocation to group
and sort all includes throught our formatted code. The grouping is as
follows:
- 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/..."`)
Because it is quite hard to understand the difference between toplevel
citus includes and toplevel postgres includes it hardcodes the list of
toplevel citus includes. In the same manner it assumes anything not
prefixed with `columnar/` or `distributed/` as a postgres include.
The sorting/grouping is enforced by CI. Since we do so with our own
script there are not changes required in our uncrustify configuration.
DESCRIPTION: Adds support for propagating `CREATE`/`DROP` database
In this PR, create and drop database support is added.
For CREATE DATABASE:
* "oid" option is not supported
* specifying "strategy" to be different than "wal_log" is not supported
* specifying "template" to be different than "template1" is not
supported
The last two are because those are not saved in `pg_database` and when
activating a node, we cannot assume what parameters were provided when
creating the database.
And "oid" is not supported because whether user specified an arbitrary
oid when creating the database is not saved in pg_database and we want
to avoid from oid collisions that might arise from attempting to use an
auto-assigned oid on workers.
Finally, in case of node activation, GRANTs for the database are also
propagated.
---------
Co-authored-by: Jelte Fennema-Nio <github-tech@jeltef.nl>
Co-authored-by: Jelte Fennema-Nio <jelte.fennema@microsoft.com>
Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>
https://app.circleci.com/pipelines/github/citusdata/citus/34550/workflows/5b802f66-2666-4623-a209-6d7799f7ee5f/jobs/1229153
```diff
VACUUM (FREEZE, PROCESS_TOAST true) local_vacuum_table;
SELECT relfrozenxid::text::integer > :frozenxid AS frozen_performed FROM pg_class
WHERE oid=:reltoastrelid::regclass;
frozen_performed
------------------
- t
+ f
(1 row)
```
Process toast option in vacuum was introduced in PG14. The failing test
was supposed to be a part of `multi_utilities.sql`, but it was included
in `pg14.sql` to avoid alternative output for PG13. See
ba62c0a148 (diff-ed03478f693155e2fe092e9ad356bf884dc097f554e8d75eff562d52bbcf7a75L255-L272)
for reference.
However, now that we don't support PG13 anymore, we can move this test
to `multi_utilities.sql`. Moving the test, plus inserting data before
running vacuum freeze such that the freeze is more meaningful and not
flaky, fixes the flakiness problem of the test.
We propagate `SECURITY LABEL [for provider] ON ROLE rolename IS
labelname` to the worker nodes.
We also make sure to run the relevant `SecLabelStmt` commands on a
newly added node by looking at roles found in `pg_shseclabel`.
See official docs for explanation on how this command works:
https://www.postgresql.org/docs/current/sql-security-label.html
This command stores the role label in the `pg_shseclabel` catalog table.
This commit also fixes the regex string in
`check_gucs_are_alphabetically_sorted.sh` script such that it escapes
the dot. Previously it was looking for all strings starting with "citus"
instead of "citus." as it should.
To test this feature, I currently make use of a special GUC to control
label provider registration in PG_init when creating the Citus extension.
While investigating replication slots leftovers
in PR https://github.com/citusdata/citus/pull/7338,
I ran into the following refactoring/cleanup
that can be done in our test suite:
- Add separate test to remove non default nodes
- Remove coordinator removal from `add_coordinator` test
Use `remove_coordinator_from_metadata` test where needed
- Don't print nodeids in `multi_multiuser_auth` and
`multi_poolinfo_usage`
tests
- Use `startswith` when checking for isolation or failure tests
- Add some dependencies accordingly in `run_test.py` for running flaky
test schedules
DESCRIPTION: Adds support from issuing role management commands from worker nodes
It's unlikely to get into a distributed deadlock with role commands, we
don't care much about them at the moment.
There were several attempts to reduce the chances of a deadlock but we
didn't any of them merged into main branch yet, see:
#7325#7016#7009
When I run this test in my local, the size of the table after the DELETE
command is around 58785792. Hence, I assume that the diffs suggest that
the Vacuum had no effect. The current solution is to run the VACUUM
command three times instead of once.
Example diff:
https://github.com/citusdata/citus/actions/runs/6722231142/attempts/1#summary-18269870674
```diff
insert into local_vacuum_table select i from generate_series(1,1000000) i;
delete from local_vacuum_table;
VACUUM local_vacuum_table;
SELECT CASE WHEN s BETWEEN 20000000 AND 25000000 THEN 22500000 ELSE s END
FROM pg_total_relation_size('local_vacuum_table') s ;
s
----------
- 22500000
+ 58785792
(1 row)
```
See more diff examples in the PR description
https://github.com/citusdata/citus/pull/7334
https://github.com/citusdata/citus/actions/runs/6745019678/attempts/1#summary-18336188930
```diff
insert into target_table SELECT a*2 FROM source_table RETURNING a;
-NOTICE: executing the command locally: SELECT bytes FROM fetch_intermediate_results(ARRAY['repartitioned_results_xxxxx_from_4213582_to_0','repartitioned_results_xxxxx_from_4213584_to_0']::text[],'localhost',57638) bytes
+NOTICE: executing the command locally: SELECT bytes FROM fetch_intermediate_results(ARRAY['repartitioned_results_3940758121873413_from_4213584_to_0','repartitioned_results_3940758121873413_from_4213582_to_0']::text[],'localhost',57638) bytes
```
The elements in the array passed to `fetch_intermediate_results` are the
same, but in the opposite order than expected.
To fix this flakiness, we can omit the `"SELECT bytes FROM
fetch_intermediate_results..."` line. From the following logs, it is
understandable that the intermediate results have been fetched.
Fix the flaky test that results in following diff by waiting until the
backend that we want to terminate really terminates, until 5secs.
```diff
--- /__w/citus/citus/src/test/regress/expected/isolation_get_all_active_transactions.out.modified 2023-11-01 16:30:57.648749795 +0000
+++ /__w/citus/citus/src/test/regress/results/isolation_get_all_active_transactions.out.modified 2023-11-01 16:30:57.656749877 +0000
@@ -114,13 +114,13 @@
--------------------
t
(1 row)
step s3-show-activity:
SET ROLE postgres;
select count(*) from get_all_active_transactions() where process_id IN (SELECT * FROM selected_pid);
count
-----
- 0
+ 1
(1 row)
```
Sometimes multi_alter_table_statements would fail in CI like this:
```diff
-- Verify that DROP NOT NULL works
ALTER TABLE lineitem_alter ALTER COLUMN int_column2 DROP NOT NULL;
SELECT "Column", "Type", "Modifiers" FROM table_desc WHERE relid='lineitem_alter'::regclass;
- Column | Type | Modifiers
----------------------------------------------------------------------
- l_orderkey | bigint | not null
- l_partkey | integer | not null
- l_suppkey | integer | not null
- l_linenumber | integer | not null
- l_quantity | numeric(15,2) | not null
- l_extendedprice | numeric(15,2) | not null
- l_discount | numeric(15,2) | not null
- l_tax | numeric(15,2) | not null
- l_returnflag | character(1) | not null
- l_linestatus | character(1) | not null
- l_shipdate | date | not null
- l_commitdate | date | not null
- l_receiptdate | date | not null
- l_shipinstruct | character(25) | not null
- l_shipmode | character(10) | not null
- l_comment | character varying(44) | not null
- float_column | double precision | default 1
- date_column | date |
- int_column1 | integer |
- int_column2 | integer |
- null_column | integer |
-(21 rows)
-
+ERROR: schema "alter_table_add_column" does not exist
-- COPY should succeed now
SELECT master_create_empty_shard('lineitem_alter') as shardid \gset
```
Reading from table_desc apparantly has an issue that if the schema gets
deleted from one of the items, while it is being read that we get such
an error.
This change fixes that by not running multi_alter_table_statements in parallel
with alter_table_add_column anymore.
This is another instance of the same issue as in #7294
Sometimes in CI we run into this failure:
```diff
SELECT resultId, nodeport, rowcount, targetShardId, targetShardIndex
FROM partition_task_list_results('test', $$ SELECT * FROM source_table $$, 'target_table')
NATURAL JOIN pg_dist_node;
-WARNING: connection to the remote node localhost:xxxxx failed with the following error: connection not open
+ERROR: connection to the remote node localhost:9060 failed with the following error: connection not open
SELECT * FROM distributed_result_info ORDER BY resultId;
- resultid | nodeport | rowcount | targetshardid | targetshardindex
----------------------------------------------------------------------
- test_from_100800_to_0 | 9060 | 22 | 100805 | 0
- test_from_100801_to_0 | 57637 | 2 | 100805 | 0
- test_from_100801_to_1 | 57637 | 15 | 100806 | 1
- test_from_100802_to_1 | 57637 | 10 | 100806 | 1
- test_from_100802_to_2 | 57637 | 5 | 100807 | 2
- test_from_100803_to_2 | 57637 | 18 | 100807 | 2
- test_from_100803_to_3 | 57637 | 4 | 100808 | 3
- test_from_100804_to_3 | 9060 | 24 | 100808 | 3
-(8 rows)
-
+ERROR: current transaction is aborted, commands ignored until end of transaction block
-- fetch from worker 2 should fail
SAVEPOINT s1;
+ERROR: current transaction is aborted, commands ignored until end of transaction block
SELECT fetch_intermediate_results('{test_from_100802_to_1,test_from_100802_to_2}'::text[], 'localhost', :worker_2_port) > 0 AS fetched;
-ERROR: could not open file "base/pgsql_job_cache/xx_x_xxx/test_from_100802_to_1.data": No such file or directory
-CONTEXT: while executing command on localhost:xxxxx
+ERROR: current transaction is aborted, commands ignored until end of transaction block
ROLLBACK TO SAVEPOINT s1;
+ERROR: savepoint "s1" does not exist
-- fetch from worker 1 should succeed
SELECT fetch_intermediate_results('{test_from_100802_to_1,test_from_100802_to_2}'::text[], 'localhost', :worker_1_port) > 0 AS fetched;
- fetched
----------------------------------------------------------------------
- t
-(1 row)
-
+ERROR: current transaction is aborted, commands ignored until end of transaction block
-- make sure the results read are same as the previous transaction block
SELECT count(*), sum(x) FROM
read_intermediate_results('{test_from_100802_to_1,test_from_100802_to_2}'::text[],'binary') AS res (x int);
- count | sum
----------------------------------------------------------------------
- 15 | 863
-(1 row)
-
+ERROR: current transaction is aborted, commands ignored until end of transaction block
ROLLBACk;
```
As outlined in the #7306 I created, the reason for this is related to
only having a single connection open to the node. Finding and fixing the
full cause is not trivial, so instead this PR starts working around
this bug by forcing maximum parallelism. Preferably we'd want
this workaround not to be necessary, but that requires
spending time to fix this. For now having a less flaky CI is
good enough.
Sometimes in CI insert_select_connection_leak would fail like this:
```diff
END;
SELECT worker_connection_count(:worker_1_port) - :pre_xact_worker_1_connections AS leaked_worker_1_connections,
worker_connection_count(:worker_2_port) - :pre_xact_worker_2_connections AS leaked_worker_2_connections;
leaked_worker_1_connections | leaked_worker_2_connections
-----------------------------+-----------------------------
- 0 | 0
+ -1 | 0
(1 row)
-- ROLLBACK
BEGIN;
INSERT INTO target_table SELECT * FROM source_table;
INSERT INTO target_table SELECT * FROM source_table;
ROLLBACK;
SELECT worker_connection_count(:worker_1_port) - :pre_xact_worker_1_connections AS leaked_worker_1_connections,
worker_connection_count(:worker_2_port) - :pre_xact_worker_2_connections AS leaked_worker_2_connections;
leaked_worker_1_connections | leaked_worker_2_connections
-----------------------------+-----------------------------
- 0 | 0
+ -1 | 0
(1 row)
\set VERBOSITY TERSE
-- Error on constraint failure
BEGIN;
INSERT INTO target_table SELECT * FROM source_table;
SELECT worker_connection_count(:worker_1_port) AS worker_1_connections,
worker_connection_count(:worker_2_port) AS worker_2_connections \gset
SAVEPOINT s1;
INSERT INTO target_table SELECT a, CASE WHEN a < 50 THEN b ELSE null END FROM source_table;
@@ -89,15 +89,15 @@
leaked_worker_1_connections | leaked_worker_2_connections
-----------------------------+-----------------------------
0 | 0
(1 row)
END;
SELECT worker_connection_count(:worker_1_port) - :pre_xact_worker_1_connections AS leaked_worker_1_connections,
worker_connection_count(:worker_2_port) - :pre_xact_worker_2_connections AS leaked_worker_2_connections;
leaked_worker_1_connections | leaked_worker_2_connections
-----------------------------+-----------------------------
- 0 | 0
+ -1 | 0
(1 row)
```
Source:
https://github.com/citusdata/citus/actions/runs/6718401194/attempts/1#summary-18258258387
A negative amount of leaked connectios is obviously not possible. For
some reason there was a connection open when we checked the initial
amount of connections that was closed afterwards. This could be the
from the maintenance daemon or maybe from the previous test that had not
fully closed its connections just yet.
The change in this PR doesnt't actually fix the cause of the negative
connection, but it simply considers it good as well, by changing the
result to zero for negative values.
With this fix we might sometimes miss a leak, because the negative
number can cancel out the leak and still result in a 0. But since the
negative number only occurs sometimes, we'll still find the leak often
enough.
When executing a prepared CALL, which is not pure SQL but available with
some drivers like npgsql and jpgdbc, Citus entered a code path where a
plan is not defined, while trying to increase its cost. Thus SIG11 when
plan is a NULL pointer.
Fix by only increasing plan cost when plan is not null.
However, it is a bit suspicious to get here with a NULL plan and maybe a
better change will be to not call
ShardPlacementForFunctionColocatedWithDistTable() with a NULL plan at
all (in call.c:134)
bug hit with for example:
```
CallableStatement proc = con.prepareCall("{CALL p(?)}");
proc.registerOutParameter(1, java.sql.Types.BIGINT);
proc.setInt(1, -100);
proc.execute();
```
where `p(bigint)` is a distributed "function" and the param the
distribution key (also in a distributed table), see #7242 for details
Fixes#7242
Sometimes in CI our logical_replication test fails like this:
```diff
+++ /__w/citus/citus/src/test/regress/results/logical_replication.out.modified 2023-11-01 14:15:08.562758546 +0000
@@ -40,21 +40,21 @@
SELECT count(*) from pg_publication;
count
-------
0
(1 row)
SELECT count(*) from pg_replication_slots;
count
-------
- 0
+ 1
(1 row)
SELECT count(*) FROM dist;
count
-------
```
It's hard to understand what is going on here, just based on the wrong
number. So this PR changes the test to show the name of the
subscription, publication and replication slot to make finding the cause
easier.
In passing this also fixes another flaky test in the same file that our
flaky test detection picked up. This is done by waiting for resource
cleanup after the shard move.
Fixes the flaky test that results in following diff:
```diff
--- /__w/citus/citus/src/test/regress/expected/multi_mx_node_metadata.out.modified 2023-11-01 14:22:12.890476575 +0000
+++ /__w/citus/citus/src/test/regress/results/multi_mx_node_metadata.out.modified 2023-11-01 14:22:12.914476657 +0000
@@ -840,24 +840,26 @@
(1 row)
\c :datname - - :master_port
SELECT datname FROM pg_stat_activity WHERE application_name LIKE 'Citus Met%';
datname
------------
db_to_drop
(1 row)
DROP DATABASE db_to_drop;
+ERROR: database "db_to_drop" is being accessed by other users
SELECT datname FROM pg_stat_activity WHERE application_name LIKE 'Citus Met%';
datname
------------
-(0 rows)
+ db_to_drop
+(1 row)
-- cleanup
DROP SEQUENCE sequence CASCADE;
NOTICE: drop cascades to default value for column a of table reference_table
```
Sometimes isolation_metadata_sync_deadlock fails in CI like this:
```diff
diff -dU10 -w /__w/citus/citus/src/test/regress/expected/isolation_metadata_sync_deadlock.out /__w/citus/citus/src/test/regress/results/isolation_metadata_sync_deadlock.out
--- /__w/citus/citus/src/test/regress/expected/isolation_metadata_sync_deadlock.out.modified 2023-11-01 16:03:15.090199229 +0000
+++ /__w/citus/citus/src/test/regress/results/isolation_metadata_sync_deadlock.out.modified 2023-11-01 16:03:15.098199312 +0000
@@ -110,10 +110,14 @@
t
(1 row)
step s2-stop-connection:
SELECT stop_session_level_connection_to_node();
stop_session_level_connection_to_node
-------------------------------------
(1 row)
+
+teardown failed: ERROR: localhost:57638 is a metadata node, but is out of sync
+HINT: If the node is up, wait until metadata gets synced to it and try again.
+CONTEXT: SQL statement "SELECT master_remove_distributed_table_metadata_from_workers(v_obj.objid, v_obj.schema_name, v_obj.object_name)"
```
Source:
https://github.com/citusdata/citus/actions/runs/6721938040/attempts/1#summary-18268946448
To fix this we now wait for the metadata to be fully synced to all
nodes at the start of the teardown steps.
Sometimes in CI citus_non_blocking_split_shard_cleanup failed like this:
```diff
--- /__w/citus/citus/src/test/regress/expected/citus_non_blocking_split_shard_cleanup.out.modified 2023-11-01 15:07:14.280551207 +0000
+++ /__w/citus/citus/src/test/regress/results/citus_non_blocking_split_shard_cleanup.out.modified 2023-11-01 15:07:14.292551358 +0000
@@ -106,21 +106,22 @@
-----------------------------------
(1 row)
\c - - - :worker_2_port
SET search_path TO "citus_split_test_schema";
-- Replication slots should be cleaned up
SELECT slot_name FROM pg_replication_slots;
slot_name
---------------------------------
-(0 rows)
+ citus_shard_split_slot_19_10_17
+(1 row)
-- Publications should be cleanedup
SELECT count(*) FROM pg_publication;
count
```
It's expected that the replication slot is sometimes not cleaned up if
we don't wait until resource cleanup completes. This PR starts doing
that here.
Sometimes failure_split_cleanup failed in CI like this:
```diff
ERROR: server closed the connection unexpectedly
CONTEXT: while executing command on localhost:9060
SELECT operation_id, object_type, object_name, node_group_id, policy_type
FROM pg_dist_cleanup where operation_id = 777 ORDER BY object_name;
operation_id | object_type | object_name | node_group_id | policy_type
--------------+-------------+-----------------------------------------------------------+---------------+-------------
777 | 1 | citus_failure_split_cleanup_schema.table_to_split_8981000 | 1 | 0
- 777 | 1 | citus_failure_split_cleanup_schema.table_to_split_8981002 | 1 | 1
777 | 1 | citus_failure_split_cleanup_schema.table_to_split_8981002 | 2 | 0
+ 777 | 1 | citus_failure_split_cleanup_schema.table_to_split_8981002 | 1 | 1
777 | 1 | citus_failure_split_cleanup_schema.table_to_split_8981003 | 2 | 1
777 | 4 | citus_shard_split_publication_1_10_777 | 2 | 0
(5 rows)
-- we need to allow connection so that we can connect to proxy
```
Source:
https://github.com/citusdata/citus/actions/runs/6717642291/attempts/1#summary-18256014949
It's the common problem where we're missing a column in the ORDER BY
clause. This fixes that by adding an node_group_id to the query in
question.
Sometimes in CI isolation_master_update_node fails like this:
```diff
------------------
(1 row)
step s2-abort: ABORT;
step s1-abort: ABORT;
FATAL: terminating connection due to administrator command
FATAL: terminating connection due to administrator command
SSL connection has been closed unexpectedly
+server closed the connection unexpectedly
master_remove_node
------------------
```
This just seesm like a random error line. The only way to reasonably fix
this is by adding an extra output file. So that's what this PR does.
One of our most flaky and most anoying tests is
multi_cluster_management. It usually fails like this:
```diff
SELECT citus_disable_node('localhost', :worker_2_port);
citus_disable_node
--------------------
(1 row)
SELECT public.wait_until_metadata_sync(60000);
+WARNING: waiting for metadata sync timed out
wait_until_metadata_sync
--------------------------
(1 row)
```
This tries to address that by hardening wait_until_metadata_sync. I
believe the reason for this warning is that there is a race condition in
wait_until_metadata_sync. It's possible for the pre-check to fail, then
have the maintenance daemon send a notification. And only then have the
backend start to listen. I tried to fix it in two ways:
1. First run LISTEN, and only then read do the pre-check.
2. If we time out, check again just to make sure that we did not miss
the notification somehow. And don't show a warning if all metadata is
synced after the timeout.
It's hard to know for sure that this fixes it because the test is not
repeatable and I could not reproduce it locally. Let's just hope for the
best.
---------
Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>
Sometimes multi_reference_table failed in CI like this:
```diff
\c - - - :master_port
DROP INDEX reference_schema.reference_index_2;
\c - - - :worker_1_port
SELECT "Column", "Type", "Modifiers" FROM table_desc WHERE relid='reference_schema.reference_table_ddl_1250019'::regclass;
- Column | Type | Modifiers
----------------------------------------------------------------------
- value_2 | double precision | default 25.0
- value_3 | text | not null
- value_4 | timestamp without time zone |
- value_5 | double precision |
-(4 rows)
-
+ERROR: schema "citus_local_table_queries" does not exist
\di reference_schema.reference_index_2*
List of relations
Schema | Name | Type | Owner | Table
```
Source:
https://github.com/citusdata/citus/actions/runs/6707535961/attempts/2#summary-18226879513
Reading from table_desc apparantly has an issue that if the schema gets
deleted from one of the items, while it is being read that we get such
an error.
This change fixes that by not running multi_reference_table in parallel
with citus_local_tables_queries anymore.
I just enhanced the existing code to check if the relation is an index
belonging to a distributed table.
If so the shardId is appended to relation (index) name and the *_size
function are executed as before.
There is a change in an extern function:
`extern StringInfo GenerateSizeQueryOnMultiplePlacements(...)`
It's possible to create a new function and deprecate this one later if
compatibility is an issue.
Fixes https://github.com/citusdata/citus/issues/6496.
DESCRIPTION: Allows using Citus size functions on distributed tables
indexes.
---------
Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>
Sometimes validate constraint would fail like this:
```diff
validatable_constraint_8000016 | t
(10 rows)
DROP TABLE constrained_table;
+ERROR: deadlock detected
+DETAIL: Process 16602 waits for ShareRowExclusiveLock on relation 56258 of database 16384; blocked by process 16601.
+Process 16601 waits for AccessShareLock on relation 56120 of database 16384; blocked by process 16602.
+HINT: See server log for query details.
DROP TABLE referenced_table CASCADE;
DROP TABLE referencing_table;
DROP SCHEMA validate_constraint CASCADE;
-NOTICE: drop cascades to 3 other objects
+NOTICE: drop cascades to 4 other objects
DETAIL: drop cascades to type constraint_validity
drop cascades to view constraint_validations_in_workers
drop cascades to view constraint_validations
+drop cascades to table constrained_table
SET search_path TO DEFAULT;
```
Source:
https://github.com/citusdata/citus/actions/runs/6708383699?pr=7291
This change fixes that by not running together with the
foreign_key_to_reference_table test anymore. In passing it also
simplifies dropping of the test its resources.
DESCRIPTION: This change starts a maintenance deamon at the time of
server start if there is a designated main database.
This is the code flow:
1. User designates a main database:
`ALTER SYSTEM SET citus.main_db = "myadmindb";`
2. When postmaster starts, in _PG_Init, citus calls
`InitializeMaintenanceDaemonForMainDb`
This function registers a background worker to run
`CitusMaintenanceDaemonMain `with `databaseOid = 0 `
3. `CitusMaintenanceDaemonMain ` takes some special actions when
databaseOid is 0:
- Gets the citus.main_db value.
- Connects to the citus.main_db
- Now the `MyDatabaseId `is available, creates a hash entry for it.
- Then follows the same control flow as for a regular db,
HasDistributionKey & HasDistributionKeyCacheEntry returns true when the
corresponding table has a distribution key, the comments state the
opposite,
which should be fixed.
Signed-off-by: Zhao Junwang <zhjwpku@gmail.com>
Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>
There was a bug reported for previous versions of Citus where
shard\_size was returning NULL for tables with spaces in them. It works
fine on the main branch though, but I'm still adding a test for this to
the main branch because it seems a good test to have.
DESCRIPTION: Send keepalive messages during the logical replication
phase of large shard splits to avoid timeouts.
During the logical replication part of the shard split process, split
decoder filters out the wal records produced by the initial copy. If the
number of wal records is big, then split decoder ends up processing for
a long time before sending out any wal records through pgoutput. Hence
the wal receiver may time out and restarts repeatedly causing our split
driver code catch up logic to fail.
Notes:
1. If the wal_receiver_timeout is set to a very small number e.g. 600ms,
it may time out before receiving the keepalives. My tests show that this
code works best when the` wal_receiver_timeout `is set to 1minute, which
is the default value.
2. Once a logical replication worker time outs, a new one gets launched.
The new logical replication worker sets the pg_stat_subscription columns
to initial values. E.g. the latest_end_lsn is set to 0. Our driver logic
in `WaitForGroupedLogicalRepTargetsToCatchUp` can not handle LSN value
to go back. This is the main reason for it to get stuck in the infinite
loop.
DESCRIPTION: Fix leaking of memory and memory contexts in Foreign
Constraint Graphs
Previously, every time we (re)created the Foreign Constraint
Relationship Graph, we created a new Memory Context while loosing a
reference to the previous context. This old context could still have
left over memory in there causing a memory leak.
With this patch we statically have one memory context that we lazily
initialize the first time we create our foreign constraint relationship
graph. On every subsequent creation, beside destroying our previous
hashmap we also reset our memory context to remove any left over
references.
This commit aims to add a comprehensive guide that covers all essential
aspects of Citus, including planning, execution, locking mechanisms,
shard moves, 2PC, and many other major components of Citus.
Co-authored-by: Marco Slot <marco.slot@gmail.com>
When testing rolling Citus upgrades, coordinator should not be upgraded
until we upgrade all the workers.
---------
Co-authored-by: Jelte Fennema-Nio <github-tech@jeltef.nl>
DESCRIPTION: Shard moves/isolate report LSN's in lsn format
While investigating an issue with our catchup mechanism on certain
postgres versions we noticed we print LSN's in the format of the native
long type. This is an uncommon representation for LSN's in postgres
logs.
This patch changes the output of our log message to go from the long
type representation to the native LSN type representation. Making it
easier for postgres users to recognize and compare LSN's with other
related reports.
example of new output:
```
2023-09-25 17:28:47.544 CEST [11345] LOG: The LSN of the target subscriptions on node localhost:9701 have increased from 0/0 to 0/E1ED20F8 at 2023-09-25 17:28:47.544165+02 where the source LSN is 1/415DCAD0
```
When cdc got added the makefiles hardcoded the `.so` extension instead
of using the platform specifc `$(DLSUFFIX)` variable used by `pgxs.mk`.
Also don't remove installed cdc artifacts on `make clean`.
This was sometimes failing when running locally due to some local shard
still existing due to. This fixes that. We normally silence all
`drop schema cascade` output like this anyway to avoid unnecessary
diffs when modifying a test later on.
DESCRIPTION: Adds support for ALTER DATABASE <db_name> SET .. statement
propagation
SET statements in Postgres has a common structure which is already being
used in Alter Function
statement.
In this PR, I added a util file; citus_setutils and made it usable for
both for
alter database<db_name>set .. and alter function ... set ... statements.
With this PR, below statements will be propagated
```sql
ALTER DATABASE name SET configuration_parameter { TO | = } { value | DEFAULT }
ALTER DATABASE name SET configuration_parameter FROM CURRENT
ALTER DATABASE name RESET configuration_parameter
ALTER DATABASE name RESET ALL
```
Additionally, there was a bug in processing float values in the common
code block.
I fixed this one as well
Previous
```C
case T_Float:
{
appendStringInfo(buf, " %s", strVal(value));
break;
}
```
Now
```C
case T_Float:
{
appendStringInfo(buf, " %s", nodeToString(value));
break;
}
```
DESCRIPTION: Adds ALTER DATABASE WITH ... and REFRESH COLLATION VERSION
support
This PR adds supports for basic ALTER DATABASE statements propagation
support. Below statements are supported:
ALTER DATABASE <database_name> with IS_TEMPLATE <true/false>;
ALTER DATABASE <database_name> with CONNECTION LIMIT <integer_value>;
ALTER DATABASE <database_name> REFRESH COLLATION VERSION;
---------
Co-authored-by: Jelte Fennema-Nio <jelte.fennema@microsoft.com>
We currently don't support propagating these options in Citus
Relevant PG commits:
https://github.com/postgres/postgres/commit/e3ce2dehttps://github.com/postgres/postgres/commit/3d14e17
Limitation:
We also need to take care of generated GRANT statements by dependencies
in attempt to distribute something else. Specifically, this part of the
code in `GenerateGrantRoleStmtsOfRole`:
```
grantRoleStmt->admin_opt = membership->admin_option;
```
In PG16, membership also has `inherit_option` and `set_option` which
need to properly be part of the `grantRoleStmt`. We can skip for now
since #7164 will take care of this soon, and also this is not an
expected use-case.
Add citus_schema_move() that can be used to move tenant tables within a distributed
schema to another node. The function has two variations as simple wrappers around
citus_move_shard_placement() and citus_move_shard_placement_with_nodeid() respectively.
They pick a shard that belongs to the given tenant schema and resolve the source node
that contain the shards under given tenant schema. Hence their signatures are quite
similar to underlying functions:
```sql
-- citus_schema_move(), using target node name and node port
CREATE OR REPLACE FUNCTION pg_catalog.citus_schema_move(
schema_id regnamespace,
target_node_name text,
target_node_port integer,
shard_transfer_mode citus.shard_transfer_mode default 'auto')
RETURNS void
LANGUAGE C STRICT
AS 'MODULE_PATHNAME', $$citus_schema_move$$;
-- citus_schema_move(), using target node id
CREATE OR REPLACE FUNCTION pg_catalog.citus_schema_move(
schema_id regnamespace,
target_node_id integer,
shard_transfer_mode citus.shard_transfer_mode default 'auto')
RETURNS void
LANGUAGE C STRICT
AS 'MODULE_PATHNAME', $$citus_schema_move_with_nodeid$$;
```
Since in PG16, truncate triggers are supported on foreign tables, we add
the citus_truncate_trigger to Citus foreign tables as well, such that the TRUNCATE
command is propagated to the table's single local shard as well.
Note that TRUNCATE command was working for foreign tables even before this
commit: see https://github.com/citusdata/citus/pull/7170#issuecomment-1706240593 for details
This commit also adds tests with user-enabled truncate triggers on Citus foreign tables:
both trigger on the shell table and on its single foreign local shard.
Relevant PG commit:
https://github.com/postgres/postgres/commit/3b00a94
**Problem:**
Previously we always used an outside superuser connection to overcome
permission issues for the current user while propagating dependencies.
That has mainly 2 problems:
1. Visibility issues during dependency propagation, (metadata connection
propagates some objects like a schema, and outside transaction does not
see it and tries to create it again)
2. Security issues (it is preferrable to use current user's connection
instead of extension superuser)
**Solution (high level):**
Now, we try to make a smarter decision on whether should we use an
outside superuser connection or current user's metadata connection. We
prefer using current user's connection if any of the objects, which is
already propagated in the current transaction, is a dependency for a
target object. We do that since we assume if current user has
permissions to create the dependency, then it can most probably
propagate the target as well.
Our assumption is expected to hold most of the times but it can still be
wrong. In those cases, transaction would fail and user should set the
GUC `citus.create_object_propagation` to `deferred` to work around it.
**Solution:**
1. We track all objects propagated in the current transaction (we can
handle subtransactions),
2. We propagate dependencies via the current user's metadata connection
if any dependency is created in the current transaction to address
issues listed above. Otherwise, we still use an outside superuser
connection.
DESCRIPTION: Fixes some object propagation errors seen with transaction
blocks.
Fixes https://github.com/citusdata/citus/issues/6614
---------
Co-authored-by: Nils Dijk <nils@citusdata.com>
For a database that does not create the citus extension by running
` CREATE EXTENSION citus;`
`CitusHasBeenLoaded ` function ends up querying the `pg_extension` table
every time it is invoked. This is not an ideal situation for a such a
database.
The idea in this PR is as follows:
### A new field in MetadataCache.
Add a new variable `extensionCreatedState `of the following type:
```
typedef enum ExtensionCreatedState
{
UNKNOWN = 0,
CREATED = 1,
NOTCREATED = 2,
} ExtensionCreatedState;
```
When the MetadataCache is invalidated, `ExtensionCreatedState` will be
set to UNKNOWN.
### Invalidate MetadataCache when CREATE/DROP/ALTER EXTENSION citus
commands are run.
- Register a callback function, named
`InvalidateDistRelationCacheCallback`, for relcache invalidation during
the shared library initialization for `citus.so`. This callback function
is invoked in all the backends whenever the relcache is invalidated in
one of the backends. (This could be caused many DDLs operations).
- In the cache invalidation callback,`
InvalidateDistRelationCacheCallback`, invalidate `MetadataCache` zeroing
it out.
- In `CitusHasBeenLoaded`, perform the costly citus is loaded check only
if the `MetadataCache` is not valid.
### Downsides
Any relcache invalidation (caused by various DDL operations) will case
Citus MetadataCache to get invalidated. Most of the time it will be
unnecessary. But we rely on that DDL operations on relations will not be
too frequent.
When braking a colocation, we need to create a new colocation group
record in pg_dist_colocation for the relation. It is not sufficient to
have a new colocationid value in pg_dist_partition only.
This patch also fixes a bug when deleting a colocation group if no
tables are left in it. Previously we passed a relation id as a parameter
to DeleteColocationGroupIfNoTablesBelong function, where we should have
passed a colocation id.
1. Adds an `sql_row` function, for when a query returns a single row
with multiple columns.
2. Include a `notice_handler` for easier debugging
3. Retry dropping replication slots when they are "in use", this is
often an ephemeral state and can cause flaky tests
In PG16, REINDEX DATABASE/SYSTEM name is optional.
We already don't propagate these commands automatically.
Testing here with run_command_on_workers.
Relevant PG commit:
https://github.com/postgres/postgres/commit/2cbc3c1
When we create a database, it already needs to be manually created in
the workers as well.
This new icu_rules option should work as the other options as well.
Added a test for that.
Relevant PG commit:
https://github.com/postgres/postgres/commit/30a53b7
DESCRIPTION: Presenting citus_pause_node UDF enabling pausing by
node_id.
citus_pause_node takes a node_id parameter and fetches all the shards in
that node and puts AccessExclusiveLock on all the shards inside that
node. With this lock, insert is disabled, until citus_pause_node
transaction is closed.
---------
Co-authored-by: Hanefi Onaldi <Hanefi.Onaldi@microsoft.com>
Jelte Fennema-Nio
Valery
zhjwpku
Karina
Onur Tirtir
Halil Ozan Akgül
Teja Mupparti
LightDB Enterprise Postgres
Gürkan İndibay
Jodi-Ann Francis
Naisila Puka
Nils Dijk
Japin Li
dependabot[bot]
cvbhjkl
Cédric Villemain
Cédric Villemain
Emel Şimşek
Benjamin O
Önder Kalacı
aykut-bozkurt
Hanefi Onaldi
Ivan Vyazmitinov