external
/
citus
mirror of https://github.com/citusdata/citus.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Add Constraint %s Add Primary Key Using index %s support 

This commit makes a change in relay_event_utility.c to check if the
Alter Table command adds a constraint using index. If this is the
case, it appends the shard id to the index name.
pull/1739/head
mehmet furkan şahin 2017-10-27 16:11:10 +03:00
parent c69d8bf4a6
commit 32fb19911c
3 changed files with 228 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
src/backend/distributed/relay/relay_event_utility.c
View File

@ -100,8 +100,19 @@ RelayEventExtendNames(Node *parseTree, char *schemaName, uint64 shardId)
{ {
AlterTableCmd *command = (AlterTableCmd *) lfirst(commandCell); AlterTableCmd *command = (AlterTableCmd *) lfirst(commandCell);
if (command->subtype == AT_AddConstraint || if (command->subtype == AT_AddConstraint)
command->subtype == AT_DropConstraint) {
Constraint *constraint = (Constraint *) command->def;
if (constraint->contype == CONSTR_PRIMARY && constraint->indexname)
{
char **indexName = &(constraint->indexname);
AppendShardIdToName(indexName, shardId);
}
AppendShardIdToConstraintName(command, shardId);
}
if (command->subtype == AT_DropConstraint)
{ {
AppendShardIdToConstraintName(command, shardId); AppendShardIdToConstraintName(command, shardId);
} }

128
src/test/regress/expected/multi_alter_table_add_constraints.out
View File

@ -488,3 +488,131 @@ SELECT "Constraint", "Definition" FROM table_checks WHERE relid='public.products
\c - - - :master_port \c - - - :master_port
DROP TABLE products; DROP TABLE products;
SET citus.shard_count to 2;
-- Test if the ALTER TABLE %s ADD %s PRIMARY KEY %s works
CREATE SCHEMA sc1;
CREATE TABLE sc1.alter_add_prim_key(x int, y int);
CREATE UNIQUE INDEX CONCURRENTLY alter_pk_idx ON sc1.alter_add_prim_key(x);
ALTER TABLE sc1.alter_add_prim_key ADD CONSTRAINT alter_pk_idx PRIMARY KEY USING INDEX alter_pk_idx;
SELECT create_distributed_table('sc1.alter_add_prim_key', 'x');
create_distributed_table
--------------------------
(1 row)
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc1' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
nodename | nodeport | success | result
-----------+----------+---------+----------------------
localhost | 57637 | t | alter_pk_idx_1450234
localhost | 57638 | t | alter_pk_idx_1450234
(2 rows)
CREATE SCHEMA sc2;
CREATE TABLE sc2.alter_add_prim_key(x int, y int);
SET search_path TO 'sc2';
SELECT create_distributed_table('alter_add_prim_key', 'x');
create_distributed_table
--------------------------
(1 row)
CREATE UNIQUE INDEX CONCURRENTLY alter_pk_idx ON alter_add_prim_key(x);
ALTER TABLE alter_add_prim_key ADD CONSTRAINT alter_pk_idx PRIMARY KEY USING INDEX alter_pk_idx;
NOTICE: using one-phase commit for distributed DDL commands
HINT: You can enable two-phase commit for extra safety with: SET citus.multi_shard_commit_protocol TO '2pc'
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc2' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
nodename | nodeport | success | result
-----------+----------+---------+----------------------
localhost | 57637 | t | alter_pk_idx_1450236
localhost | 57638 | t | alter_pk_idx_1450236
(2 rows)
-- We are running almost the same test with a slight change on the constraint name because if the constraint has a different name than the index, Postgres renames the index.
CREATE SCHEMA sc3;
CREATE TABLE sc3.alter_add_prim_key(x int);
INSERT INTO sc3.alter_add_prim_key(x) SELECT generate_series(1,100);
SET search_path TO 'sc3';
SELECT create_distributed_table('alter_add_prim_key', 'x');
NOTICE: Copying data from local table...
create_distributed_table
--------------------------
(1 row)
CREATE UNIQUE INDEX CONCURRENTLY alter_pk_idx ON alter_add_prim_key(x);
ALTER TABLE alter_add_prim_key ADD CONSTRAINT a_constraint PRIMARY KEY USING INDEX alter_pk_idx;
NOTICE: ALTER TABLE / ADD CONSTRAINT USING INDEX will rename index "alter_pk_idx" to "a_constraint"
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc3' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
nodename | nodeport | success | result
-----------+----------+---------+----------------------
localhost | 57637 | t | a_constraint_1450238
localhost | 57638 | t | a_constraint_1450238
(2 rows)
ALTER TABLE alter_add_prim_key DROP CONSTRAINT a_constraint;
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc3' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
nodename | nodeport | success | result
-----------+----------+---------+--------
localhost | 57637 | t |
localhost | 57638 | t |
(2 rows)
SET search_path TO 'public';
DROP SCHEMA sc1 CASCADE;
NOTICE: drop cascades to table sc1.alter_add_prim_key
DROP SCHEMA sc2 CASCADE;
NOTICE: drop cascades to table sc2.alter_add_prim_key
DROP SCHEMA sc3 CASCADE;
NOTICE: drop cascades to table sc3.alter_add_prim_key

87
src/test/regress/sql/multi_alter_table_add_constraints.sql
View File

@ -422,3 +422,90 @@ SELECT "Constraint", "Definition" FROM table_checks WHERE relid='public.products
\c - - - :master_port \c - - - :master_port
DROP TABLE products; DROP TABLE products;
SET citus.shard_count to 2;
-- Test if the ALTER TABLE %s ADD %s PRIMARY KEY %s works
CREATE SCHEMA sc1;
CREATE TABLE sc1.alter_add_prim_key(x int, y int);
CREATE UNIQUE INDEX CONCURRENTLY alter_pk_idx ON sc1.alter_add_prim_key(x);
ALTER TABLE sc1.alter_add_prim_key ADD CONSTRAINT alter_pk_idx PRIMARY KEY USING INDEX alter_pk_idx;
SELECT create_distributed_table('sc1.alter_add_prim_key', 'x');
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc1' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
CREATE SCHEMA sc2;
CREATE TABLE sc2.alter_add_prim_key(x int, y int);
SET search_path TO 'sc2';
SELECT create_distributed_table('alter_add_prim_key', 'x');
CREATE UNIQUE INDEX CONCURRENTLY alter_pk_idx ON alter_add_prim_key(x);
ALTER TABLE alter_add_prim_key ADD CONSTRAINT alter_pk_idx PRIMARY KEY USING INDEX alter_pk_idx;
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc2' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
-- We are running almost the same test with a slight change on the constraint name because if the constraint has a different name than the index, Postgres renames the index.
CREATE SCHEMA sc3;
CREATE TABLE sc3.alter_add_prim_key(x int);
INSERT INTO sc3.alter_add_prim_key(x) SELECT generate_series(1,100);
SET search_path TO 'sc3';
SELECT create_distributed_table('alter_add_prim_key', 'x');
CREATE UNIQUE INDEX CONCURRENTLY alter_pk_idx ON alter_add_prim_key(x);
ALTER TABLE alter_add_prim_key ADD CONSTRAINT a_constraint PRIMARY KEY USING INDEX alter_pk_idx;
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc3' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
ALTER TABLE alter_add_prim_key DROP CONSTRAINT a_constraint;
SELECT (run_command_on_workers($$
SELECT
kc.constraint_name
FROM
information_schema.table_constraints tc join information_schema.key_column_usage kc on (kc.table_name = tc.table_name and kc.table_schema = tc.table_schema and kc.constraint_name = tc.constraint_name)
WHERE
kc.table_schema = 'sc3' and tc.constraint_type = 'PRIMARY KEY' and kc.table_name LIKE 'alter_add_prim_key_%'
ORDER BY
1
LIMIT
1;
$$)).*
ORDER BY
1,2,3,4;
SET search_path TO 'public';
DROP SCHEMA sc1 CASCADE;
DROP SCHEMA sc2 CASCADE;
DROP SCHEMA sc3 CASCADE;