Add support for creating distributed tables with a null shard key #6745

onurctirtir · 2023-03-03T08:25:48Z

With this PR, we allow creating distributed tables with without
specifying a shard key via create_distributed_table(). Here are the
the important details about those tables:

Specifying shard_count is not allowed because it is assumed to be 1.
We mostly call such tables as "null shard-key" table in code / comments.
To avoid doing a breaking layout change in create_distributed_table();
instead of throwing an error, it will inform the user that distribution_type
param is ignored unless it's explicitly set to NULL.
colocate_with param allows colocating such null shard-key tables to
each other.
We define this table type, i.e., NULL_SHARD_KEY_TABLE, as a subclass of
DISTRIBUTED_TABLE because we mostly want to treat them as distributed
tables in terms of SQL / DDL / operation support.
Metadata for such tables look like:
- distribution method => DISTRIBUTE_BY_NONE
- replication model => REPLICATION_MODEL_STREAMING
- colocation id => != INVALID_COLOCATION_ID (distinguishes from Citus local tables)
We assign colocation groups for such tables to different nodes in a
round-robin fashion based on the modulo of "colocation id".

Note that this PR doesn't care about DDL (except CREATE TABLE) / SQL /
operation (i.e., Citus UDFs) support for such tables but adds a preliminary
API.

codecov · 2023-03-03T08:36:17Z

Codecov Report

Merging #6745 (7b94697) into main (cf55136) will decrease coverage by 1.12%.
The diff coverage is 92.66%.

❗ Current head 7b94697 differs from pull request most recent head e16df14. Consider uploading reports for the commit e16df14 to get more accurate results

@@            Coverage Diff             @@
##             main    #6745      +/-   ##
==========================================
- Coverage   93.17%   92.05%   -1.12%     
==========================================
  Files         260      259       -1     
  Lines       56106    56031      -75     
==========================================
- Hits        52275    51579     -696     
- Misses       3831     4452     +621

src/backend/distributed/commands/create_distributed_table.c

onurctirtir · 2023-03-03T08:43:24Z

src/backend/distributed/operations/create_shards.c

+	if (targetCacheEntry->partitionMethod != DISTRIBUTE_BY_NONE)
+	{
+		/* make sure that tables are hash partitioned */
+		CheckHashPartitionedTable(targetRelationId);
+		CheckHashPartitionedTable(sourceRelationId);
+	}


even more, we can drop this check because at some point we anyway check whether two tables can be colocated (in ColocationIdForNewTable())

I suppose range-partitioned tables can be co-located

so do you think we shouldn't have had those checks already?

maybe change to an assert

On main, this is what happens if try using colocate_with option for range distributed tables:

SELECT create_distributed_table('tbl_1', 'a', distribution_type=>'range', colocate_with=>'tbl'); ERROR: cannot distribute relation DETAIL: Currently, colocate_with option is only supported for hash distributed tables.

So I think removing those checks / converting them to Assert()s wouldn't look like something that this PR is expected to do, what do you think?

that's not this error though, can we really get here? it does not appear in tests

I think those are more like sanity checks, so let me simply make them asserts then.

Oh we can't make them asserts because their return type is void, so let me leave them as is.

Assert(targetCacheEntry->partitionMethod == DISTRIBUTE_BY_HASH || targetCacheEntry->partitionMethod == DISTRIBUTE_BY_NONE);

src/backend/distributed/commands/create_distributed_table.c

src/backend/distributed/utils/colocation_utils.c

marcocitus · 2023-03-10T10:42:20Z

src/backend/distributed/operations/create_shards.c

+	}
+
+	static int roundRobinNodeIdx = -1;
+	roundRobinNodeIdx++;


This seems local to the session. Is there a reliable way to make it round robin across sessions? (e.g. using shard ID)

marcocitus · 2023-03-10T10:43:05Z

src/backend/distributed/operations/create_shards.c

+	LockRelationOid(DistNodeRelationId(), RowShareLock);
+
+	/* load and sort the worker node list for deterministic placement */
+	List *workerNodeList = DistributedTablePlacementNodeList(NoLock);


seems like we could use RowShareLock and omit the line above

marcocitus · 2023-03-10T10:45:33Z

src/backend/distributed/commands/create_distributed_table.c

+	}
+	else
+	{
+		CreateNullKeyShard(relationId, useExclusiveConnection);


wonder if we should add something like RoundRobin to the name to indiciate that this function is not to be used for co-located null key table creation

marcocitus

Overall this looks reasonable. Do we want to merge soon and add tests in subsequent PRs or is this the beginning of a feature branch? (I think I'm fine either way)

onurctirtir · 2023-03-13T09:14:52Z

src/test/regress/expected/create_null_dist_key.out

+(2 rows)
+
+-- This should fail due to missing OVERRIDING SYSTEM VALUE
+-- after merging https://github.com/citusdata/citus/pull/6738.


(see #6738 (comment))

onurctirtir · 2023-03-15T11:50:40Z

crash:

create table s1 (x int, y int);
create table s2 (x int, y int);
create table s3 (x int, y int);
create table s4 (x int, y int);
select create_distributed_table('s1');
select create_distributed_table('s2');
select create_distributed_table('s3', NULL, colocate_with := 'none');
select create_distributed_table('s4', NULL, colocate_with := 's3');
insert into s2 select * from s3;

stack trace:

#0  0x00007f9831fafa25 in InsertPartitionColumnMatchesSelect (query=query@entry=0x564ba1740448, insertRte=insertRte@entry=0x564ba14e0fe8, 
    subqueryRte=subqueryRte@entry=0x564ba1740930, selectPartitionColumnTableId=selectPartitionColumnTableId@entry=0x7fff9c820664) at planner/insert_select_planner.c:1285
1285                    if (originalAttrNo != insertPartitionColumn->varattno)
#0  0x00007f9831fafa25 in InsertPartitionColumnMatchesSelect (query=query@entry=0x564ba1740448, insertRte=insertRte@entry=0x564ba14e0fe8, 
    subqueryRte=subqueryRte@entry=0x564ba1740930, selectPartitionColumnTableId=selectPartitionColumnTableId@entry=0x7fff9c820664) at planner/insert_select_planner.c:1285
#1  0x00007f9831fb007b in DistributedInsertSelectSupported (queryTree=queryTree@entry=0x564ba1740448, insertRte=insertRte@entry=0x564ba14e0fe8, 
    subqueryRte=subqueryRte@entry=0x564ba1740930, allReferenceTables=allReferenceTables@entry=false) at planner/insert_select_planner.c:721
#2  0x00007f9831fb0ad9 in CreateDistributedInsertSelectPlan (originalQuery=originalQuery@entry=0x564ba1740448, 
    plannerRestrictionContext=plannerRestrictionContext@entry=0x564ba16b1d18) at planner/insert_select_planner.c:292
#3  0x00007f9831fb0f5e in CreateInsertSelectPlanInternal (planId=planId@entry=1, originalQuery=originalQuery@entry=0x564ba1740448, 
    plannerRestrictionContext=plannerRestrictionContext@entry=0x564ba16b1d18, boundParams=boundParams@entry=0x0) at planner/insert_select_planner.c:244
#4  0x00007f9831fb104c in CreateInsertSelectPlan (planId=planId@entry=1, originalQuery=originalQuery@entry=0x564ba1740448, 
    plannerRestrictionContext=plannerRestrictionContext@entry=0x564ba16b1d18, boundParams=boundParams@entry=0x0) at planner/insert_select_planner.c:211
#5  0x00007f9831faaa91 in CreateDistributedPlan (planId=planId@entry=1, allowRecursivePlanning=allowRecursivePlanning@entry=true, originalQuery=0x564ba1740448, 
    query=0x564ba14e1100, boundParams=0x0, hasUnresolvedParams=hasUnresolvedParams@entry=false, plannerRestrictionContext=0x564ba16b1d18) at planner/distributed_planner.c:995
#6  0x00007f9831fab68d in CreateDistributedPlannedStmt (planContext=planContext@entry=0x7fff9c820970) at planner/distributed_planner.c:794
#7  0x00007f9831fabaa9 in PlanDistributedStmt (planContext=planContext@entry=0x7fff9c820970, rteIdCounter=<optimized out>, rteIdCounter@entry=3)
    at planner/distributed_planner.c:730
#8  0x00007f9831fac536 in distributed_planner (parse=0x564ba14e1100, query_string=<optimized out>, cursorOptions=<optimized out>, boundParams=0x0)
    at planner/distributed_planner.c:264
#9  0x00007f9830d1f0e3 in PgmongoPlanner (parse=0x564ba14e1100, queryString=0x564ba14e0100 "insert into s2 select * from s3;", cursorOptions=2048, boundParams=0x0)
    at src/planner/pgmongo_planner.c:111
#10 0x00007f9831eac79d in pgss_planner (parse=0x564ba14e1100, query_string=0x564ba14e0100 "insert into s2 select * from s3;", cursorOptions=2048, boundParams=0x0)
    at pg_stat_statements.c:953
#11 0x0000564ba0756b3a in planner (parse=parse@entry=0x564ba14e1100, query_string=query_string@entry=0x564ba14e0100 "insert into s2 select * from s3;", 
    cursorOptions=cursorOptions@entry=2048, boundParams=boundParams@entry=0x0) at planner.c:275
#12 0x0000564ba084268c in pg_plan_query (querytree=querytree@entry=0x564ba14e1100, query_string=query_string@entry=0x564ba14e0100 "insert into s2 select * from s3;", 
    cursorOptions=cursorOptions@entry=2048, boundParams=boundParams@entry=0x0) at postgres.c:883
#13 0x0000564ba0842768 in pg_plan_queries (querytrees=0x564ba1740398, query_string=query_string@entry=0x564ba14e0100 "insert into s2 select * from s3;", 
    cursorOptions=cursorOptions@entry=2048, boundParams=boundParams@entry=0x0) at postgres.c:975
#14 0x0000564ba0842c99 in exec_simple_query (query_string=query_string@entry=0x564ba14e0100 "insert into s2 select * from s3;") at postgres.c:1169
#15 0x0000564ba0844d14 in PostgresMain (dbname=<optimized out>, username=<optimized out>) at postgres.c:4593
#16 0x0000564ba079d456 in BackendRun (port=port@entry=0x564ba1581250) at postmaster.c:4511
#17 0x0000564ba07a064b in BackendStartup (port=port@entry=0x564ba1581250) at postmaster.c:4239
#18 0x0000564ba07a0894 in ServerLoop () at postmaster.c:1806
#19 0x0000564ba07a1e66 in PostmasterMain (argc=3, argv=<optimized out>) at postmaster.c:1478
#20 0x0000564ba06e289c in main (argc=3, argv=0x564ba14d8d00) at main.c:202

Added a temporary check to prevent crashing: https://github.com/citusdata/citus/blob/create-null-shard-key-tbl/src/backend/distributed/planner/insert_select_planner.c#L718-L724

onurctirtir · 2023-03-15T11:52:43Z

fwiw, this errors nicely:

CREATE OR REPLACE FUNCTION test(x int)
RETURNS void
AS $$
DECLARE
BEGIN
    RAISE NOTICE 'x %', x;
END;
$$ LANGUAGE plpgsql;
SELECT create_distributed_function('test(int)', 'x', colocate_with :='s4');
ERROR:  cannot colocate function "test" and table "s4" because colocate_with option is only supported for hash distributed tables and reference tables.

(we should think a bit about function call delegation for single shard tables, but not right now)

Pushed some changes to throw an error when the shard key is NULL, though we might want to add support for create_distributed_table_concurrently() later on.

marcocitus · 2023-03-17T08:34:48Z

apart from 2 latest comments, this looks pretty good to me

) With this PR, we allow creating distributed tables with without specifying a shard key via create_distributed_table(). Here are the the important details about those tables: * Specifying `shard_count` is not allowed because it is assumed to be 1. * We mostly call such tables as "null shard-key" table in code / comments. * To avoid doing a breaking layout change in create_distributed_table(); instead of throwing an error, it will inform the user that `distribution_type` param is ignored unless it's explicitly set to NULL or 'h'. * `colocate_with` param allows colocating such null shard-key tables to each other. * We define this table type, i.e., NULL_SHARD_KEY_TABLE, as a subclass of DISTRIBUTED_TABLE because we mostly want to treat them as distributed tables in terms of SQL / DDL / operation support. * Metadata for such tables look like: - distribution method => DISTRIBUTE_BY_NONE - replication model => REPLICATION_MODEL_STREAMING - colocation id => **!=** INVALID_COLOCATION_ID (distinguishes from Citus local tables) * We assign colocation groups for such tables to different nodes in a round-robin fashion based on the modulo of "colocation id". Note that this PR doesn't care about DDL (except CREATE TABLE) / SQL / operation (i.e., Citus UDFs) support for such tables but adds a preliminary API.

In this release, I tried something different. I experimented with adding the PR number and title to the changelog right before each changelog entry. This way, it is easier to track where a particular changelog entry comes from. After reviews are over, I plan to remove those lines with PR numbers and titles. I went through all the PRs that are merged after 11.2.0 release and came up with a list of PRs that may need help with changelog entries. You can see details on PRs grouped in several sections below. ## PRs with missing entries The following PRs below do not have a changelog entry. If you think that this is a mistake, please share it in this PR along with a suggestion on what the changelog item should be. PR #6846 : fix 3 flaky tests in failure schedule PR #6844 : Add CPU usage to citus_stat_tenants PR #6833 : Fix citus_stat_tenants period updating bug PR #6787 : Add more tests for ddl coverage PR #6842 : Add build-cdc-* temporary directories to .gitignore PR #6841 : Add build-cdc-* temporary directories to .gitignore PR #6840 : Bump Citus to 12.0devel PR #6824 : Fixes flakiness in multi_metadata_sync test PR #6811 : Backport identity column improvements to v11.2 PR #6830 : In run_test.py actually return worker_count PR #6825 : Fixes flakiness in multi_cluster_management test PR #6816 : Refactor run_test.py PR #6817 : Explicitly disallow local rels when inserting into dist table PR #6821 : Rename citus stats tenants PR #6822 : Add some more tests for initial sql support PR #6819 : Fix flakyness in citus_split_shard_by_split_points_deferred_drop PR #6814 : Make python-regress based tests runnable with run_test.py PR #6813 : Fix flaky multi_mx_schema_support test PR #6720 : Convert columnar tap tests to pytest PR #6812 : Revoke statistics permissions from public and grant them to pg_monitor PR #6769 : Citus stats tenants guc PR #6807 : Fix the incorrect (constant) value passed to pointer-to-bool parameter, pass a NULL as the value is not used PR #6797 : Attribute local queries and cached plans on local execution PR #6796 : Parse the annotation string correctly PR #6762 : Add logs to citus_stats_tenants PR #6773 : Add initial sql support for distributed tables that don't have a shard key PR #6792 : Disentangle MERGE planning code from the modify-planning code path PR #6761 : Citus stats tenants collector view PR #6791 : Make 8 more tests runnable multiple times via run_test.py PR #6786 : Refactor some of the planning code to accommodate a new planning path for MERGE SQL PR #6789 : Rename AllRelations.. functions to AllDistributedRelations.. PR #6788 : Actually skip arbitrary_configs_router & nested_execution for AllNullDistKeyDefaultConfig PR #6783 : Add a config for arbitrary config tests where all the tables are null-shard-key tables PR #6784 : Fix attach partition: citus local to null distributed PR #6782 : Add an arbitrary config test heavily based on multi_router_planner_fast_path.sql PR #6781 : Decide what to do with router planner error at one place PR #6778 : Support partitioning for dist tables with null dist keys PR #6766 : fix pip lock file PR #6764 : Make workerCount configurable for regression tests PR #6745 : Add support for creating distributed tables with a null shard key PR #6696 : This implements MERGE phase-III PR #6767 : Add pytest depedencies to Pipfile PR #6760 : Decide core distribution params in CreateCitusTable PR #6759 : Add multi_create_fdw into minimal_schedule PR #6743 : Replace CITUS_TABLE_WITH_NO_DIST_KEY checks with HasDistributionKey() PR #6751 : Stabilize single_node.sql and others that report illegal node removal PR #6742 : Refactor CreateDistributedTable() PR #6747 : Remove unused lock functions PR #6744 : Fix multiple output version arbitrary config tests PR #6741 : Stabilize single node tests PR #6740 : Fix string eval bug in migration files check PR #6736 : Make run_test.py and create_test.py importable without errors PR #6734 : Don't blanket ignore flake8 E402 error PR #6737 : Fixes bookworm packaging pipeline problem PR #6735 : Fix run_test.py on python 3.9 PR #6733 : MERGE: In deparser, add missing check for RETURNING clause. PR #6714 : Remove auto_explain workaround in citus explain hook for ALTER TABLE PR #6719 : Fix flaky test PR #6718 : Add more powerfull dependency tracking to run_test.py PR #6710 : Install non-vulnerable cryptography package PR #6711 : Support compilation and run tests on latest PG versions PR #6700 : Add auto-formatting and linting to our python code PR #6707 : Allow multi_insert_select to run repeatably PR #6708 : Fix flakyness in failure_create_distributed_table_non_empty PR #6698 : Miscellaneous cleanup PR #6704 : Update README for 11.2 PR #6703 : Fix dubious ownership error from git PR #6690 : Bump Citus to 11.3devel ## Too long changelog entries The following PRs have changelog entries that are too long to fit in a single line. I'd expect authors to supply at changelog entries in `DESCRIPTION:` lines that are at most 78 characters. If you want to supply multi-line changelog items, you can have multiple lines that start with `DESCRIPTION:` instead. PR #6837 : fixes update propagation bug when `citus_set_coordinator_host` is called more than once PR #6738 : Identity column implementation refactorings PR #6756 : Schedule parallel shard moves in background rebalancer by removing task dependencies between shard moves across colocation groups. PR #6793 : Add a GUC to disallow planning the queries that reference non-colocated tables via router planner PR #6726 : fix memory leak during altering distributed table with a lot of partition and shards PR #6722 : fix memory leak during distribution of a table with a lot of partitions PR #6693 : prevent memory leak during ConvertTable with a lot of partitions ## Empty changelog entries. The following PR had an empty `DESCRIPTION:` line. This generates an empty changelog line that needs to be removed manually. Please either provide a short entry, or remove `DESCRIPTION:` line completely. PR #6810 : Make CDC decoder an independent extension PR #6827 : Makefile changes to build CDC in builddir for pgoutput and wal2json. --------- Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>

In this release, I tried something different. I experimented with adding the PR number and title to the changelog right before each changelog entry. This way, it is easier to track where a particular changelog entry comes from. After reviews are over, I plan to remove those lines with PR numbers and titles. I went through all the PRs that are merged after 11.2.0 release and came up with a list of PRs that may need help with changelog entries. You can see details on PRs grouped in several sections below. The following PRs below do not have a changelog entry. If you think that this is a mistake, please share it in this PR along with a suggestion on what the changelog item should be. PR #6846 : fix 3 flaky tests in failure schedule PR #6844 : Add CPU usage to citus_stat_tenants PR #6833 : Fix citus_stat_tenants period updating bug PR #6787 : Add more tests for ddl coverage PR #6842 : Add build-cdc-* temporary directories to .gitignore PR #6841 : Add build-cdc-* temporary directories to .gitignore PR #6840 : Bump Citus to 12.0devel PR #6824 : Fixes flakiness in multi_metadata_sync test PR #6811 : Backport identity column improvements to v11.2 PR #6830 : In run_test.py actually return worker_count PR #6825 : Fixes flakiness in multi_cluster_management test PR #6816 : Refactor run_test.py PR #6817 : Explicitly disallow local rels when inserting into dist table PR #6821 : Rename citus stats tenants PR #6822 : Add some more tests for initial sql support PR #6819 : Fix flakyness in citus_split_shard_by_split_points_deferred_drop PR #6814 : Make python-regress based tests runnable with run_test.py PR #6813 : Fix flaky multi_mx_schema_support test PR #6720 : Convert columnar tap tests to pytest PR #6812 : Revoke statistics permissions from public and grant them to pg_monitor PR #6769 : Citus stats tenants guc PR #6807 : Fix the incorrect (constant) value passed to pointer-to-bool parameter, pass a NULL as the value is not used PR #6797 : Attribute local queries and cached plans on local execution PR #6796 : Parse the annotation string correctly PR #6762 : Add logs to citus_stats_tenants PR #6773 : Add initial sql support for distributed tables that don't have a shard key PR #6792 : Disentangle MERGE planning code from the modify-planning code path PR #6761 : Citus stats tenants collector view PR #6791 : Make 8 more tests runnable multiple times via run_test.py PR #6786 : Refactor some of the planning code to accommodate a new planning path for MERGE SQL PR #6789 : Rename AllRelations.. functions to AllDistributedRelations.. PR #6788 : Actually skip arbitrary_configs_router & nested_execution for AllNullDistKeyDefaultConfig PR #6783 : Add a config for arbitrary config tests where all the tables are null-shard-key tables PR #6784 : Fix attach partition: citus local to null distributed PR #6782 : Add an arbitrary config test heavily based on multi_router_planner_fast_path.sql PR #6781 : Decide what to do with router planner error at one place PR #6778 : Support partitioning for dist tables with null dist keys PR #6766 : fix pip lock file PR #6764 : Make workerCount configurable for regression tests PR #6745 : Add support for creating distributed tables with a null shard key PR #6696 : This implements MERGE phase-III PR #6767 : Add pytest depedencies to Pipfile PR #6760 : Decide core distribution params in CreateCitusTable PR #6759 : Add multi_create_fdw into minimal_schedule PR #6743 : Replace CITUS_TABLE_WITH_NO_DIST_KEY checks with HasDistributionKey() PR #6751 : Stabilize single_node.sql and others that report illegal node removal PR #6742 : Refactor CreateDistributedTable() PR #6747 : Remove unused lock functions PR #6744 : Fix multiple output version arbitrary config tests PR #6741 : Stabilize single node tests PR #6740 : Fix string eval bug in migration files check PR #6736 : Make run_test.py and create_test.py importable without errors PR #6734 : Don't blanket ignore flake8 E402 error PR #6737 : Fixes bookworm packaging pipeline problem PR #6735 : Fix run_test.py on python 3.9 PR #6733 : MERGE: In deparser, add missing check for RETURNING clause. PR #6714 : Remove auto_explain workaround in citus explain hook for ALTER TABLE PR #6719 : Fix flaky test PR #6718 : Add more powerfull dependency tracking to run_test.py PR #6710 : Install non-vulnerable cryptography package PR #6711 : Support compilation and run tests on latest PG versions PR #6700 : Add auto-formatting and linting to our python code PR #6707 : Allow multi_insert_select to run repeatably PR #6708 : Fix flakyness in failure_create_distributed_table_non_empty PR #6698 : Miscellaneous cleanup PR #6704 : Update README for 11.2 PR #6703 : Fix dubious ownership error from git PR #6690 : Bump Citus to 11.3devel The following PRs have changelog entries that are too long to fit in a single line. I'd expect authors to supply at changelog entries in `DESCRIPTION:` lines that are at most 78 characters. If you want to supply multi-line changelog items, you can have multiple lines that start with `DESCRIPTION:` instead. PR #6837 : fixes update propagation bug when `citus_set_coordinator_host` is called more than once PR #6738 : Identity column implementation refactorings PR #6756 : Schedule parallel shard moves in background rebalancer by removing task dependencies between shard moves across colocation groups. PR #6793 : Add a GUC to disallow planning the queries that reference non-colocated tables via router planner PR #6726 : fix memory leak during altering distributed table with a lot of partition and shards PR #6722 : fix memory leak during distribution of a table with a lot of partitions PR #6693 : prevent memory leak during ConvertTable with a lot of partitions The following PR had an empty `DESCRIPTION:` line. This generates an empty changelog line that needs to be removed manually. Please either provide a short entry, or remove `DESCRIPTION:` line completely. PR #6810 : Make CDC decoder an independent extension PR #6827 : Makefile changes to build CDC in builddir for pgoutput and wal2json. --------- Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com> (cherry picked from commit 9344300)

) With this PR, we allow creating distributed tables with without specifying a shard key via create_distributed_table(). Here are the the important details about those tables: * Specifying `shard_count` is not allowed because it is assumed to be 1. * We mostly call such tables as "null shard-key" table in code / comments. * To avoid doing a breaking layout change in create_distributed_table(); instead of throwing an error, it will inform the user that `distribution_type` param is ignored unless it's explicitly set to NULL or 'h'. * `colocate_with` param allows colocating such null shard-key tables to each other. * We define this table type, i.e., NULL_SHARD_KEY_TABLE, as a subclass of DISTRIBUTED_TABLE because we mostly want to treat them as distributed tables in terms of SQL / DDL / operation support. * Metadata for such tables look like: - distribution method => DISTRIBUTE_BY_NONE - replication model => REPLICATION_MODEL_STREAMING - colocation id => **!=** INVALID_COLOCATION_ID (distinguishes from Citus local tables) * We assign colocation groups for such tables to different nodes in a round-robin fashion based on the modulo of "colocation id". Note that this PR doesn't care about DDL (except CREATE TABLE) / SQL / operation (i.e., Citus UDFs) support for such tables but adds a preliminary API.

In this release, I tried something different. I experimented with adding the PR number and title to the changelog right before each changelog entry. This way, it is easier to track where a particular changelog entry comes from. After reviews are over, I plan to remove those lines with PR numbers and titles. I went through all the PRs that are merged after 11.2.0 release and came up with a list of PRs that may need help with changelog entries. You can see details on PRs grouped in several sections below. ## PRs with missing entries The following PRs below do not have a changelog entry. If you think that this is a mistake, please share it in this PR along with a suggestion on what the changelog item should be. PR #6846 : fix 3 flaky tests in failure schedule PR #6844 : Add CPU usage to citus_stat_tenants PR #6833 : Fix citus_stat_tenants period updating bug PR #6787 : Add more tests for ddl coverage PR #6842 : Add build-cdc-* temporary directories to .gitignore PR #6841 : Add build-cdc-* temporary directories to .gitignore PR #6840 : Bump Citus to 12.0devel PR #6824 : Fixes flakiness in multi_metadata_sync test PR #6811 : Backport identity column improvements to v11.2 PR #6830 : In run_test.py actually return worker_count PR #6825 : Fixes flakiness in multi_cluster_management test PR #6816 : Refactor run_test.py PR #6817 : Explicitly disallow local rels when inserting into dist table PR #6821 : Rename citus stats tenants PR #6822 : Add some more tests for initial sql support PR #6819 : Fix flakyness in citus_split_shard_by_split_points_deferred_drop PR #6814 : Make python-regress based tests runnable with run_test.py PR #6813 : Fix flaky multi_mx_schema_support test PR #6720 : Convert columnar tap tests to pytest PR #6812 : Revoke statistics permissions from public and grant them to pg_monitor PR #6769 : Citus stats tenants guc PR #6807 : Fix the incorrect (constant) value passed to pointer-to-bool parameter, pass a NULL as the value is not used PR #6797 : Attribute local queries and cached plans on local execution PR #6796 : Parse the annotation string correctly PR #6762 : Add logs to citus_stats_tenants PR #6773 : Add initial sql support for distributed tables that don't have a shard key PR #6792 : Disentangle MERGE planning code from the modify-planning code path PR #6761 : Citus stats tenants collector view PR #6791 : Make 8 more tests runnable multiple times via run_test.py PR #6786 : Refactor some of the planning code to accommodate a new planning path for MERGE SQL PR #6789 : Rename AllRelations.. functions to AllDistributedRelations.. PR #6788 : Actually skip arbitrary_configs_router & nested_execution for AllNullDistKeyDefaultConfig PR #6783 : Add a config for arbitrary config tests where all the tables are null-shard-key tables PR #6784 : Fix attach partition: citus local to null distributed PR #6782 : Add an arbitrary config test heavily based on multi_router_planner_fast_path.sql PR #6781 : Decide what to do with router planner error at one place PR #6778 : Support partitioning for dist tables with null dist keys PR #6766 : fix pip lock file PR #6764 : Make workerCount configurable for regression tests PR #6745 : Add support for creating distributed tables with a null shard key PR #6696 : This implements MERGE phase-III PR #6767 : Add pytest depedencies to Pipfile PR #6760 : Decide core distribution params in CreateCitusTable PR #6759 : Add multi_create_fdw into minimal_schedule PR #6743 : Replace CITUS_TABLE_WITH_NO_DIST_KEY checks with HasDistributionKey() PR #6751 : Stabilize single_node.sql and others that report illegal node removal PR #6742 : Refactor CreateDistributedTable() PR #6747 : Remove unused lock functions PR #6744 : Fix multiple output version arbitrary config tests PR #6741 : Stabilize single node tests PR #6740 : Fix string eval bug in migration files check PR #6736 : Make run_test.py and create_test.py importable without errors PR #6734 : Don't blanket ignore flake8 E402 error PR #6737 : Fixes bookworm packaging pipeline problem PR #6735 : Fix run_test.py on python 3.9 PR #6733 : MERGE: In deparser, add missing check for RETURNING clause. PR #6714 : Remove auto_explain workaround in citus explain hook for ALTER TABLE PR #6719 : Fix flaky test PR #6718 : Add more powerfull dependency tracking to run_test.py PR #6710 : Install non-vulnerable cryptography package PR #6711 : Support compilation and run tests on latest PG versions PR #6700 : Add auto-formatting and linting to our python code PR #6707 : Allow multi_insert_select to run repeatably PR #6708 : Fix flakyness in failure_create_distributed_table_non_empty PR #6698 : Miscellaneous cleanup PR #6704 : Update README for 11.2 PR #6703 : Fix dubious ownership error from git PR #6690 : Bump Citus to 11.3devel ## Too long changelog entries The following PRs have changelog entries that are too long to fit in a single line. I'd expect authors to supply at changelog entries in `DESCRIPTION:` lines that are at most 78 characters. If you want to supply multi-line changelog items, you can have multiple lines that start with `DESCRIPTION:` instead. PR #6837 : fixes update propagation bug when `citus_set_coordinator_host` is called more than once PR #6738 : Identity column implementation refactorings PR #6756 : Schedule parallel shard moves in background rebalancer by removing task dependencies between shard moves across colocation groups. PR #6793 : Add a GUC to disallow planning the queries that reference non-colocated tables via router planner PR #6726 : fix memory leak during altering distributed table with a lot of partition and shards PR #6722 : fix memory leak during distribution of a table with a lot of partitions PR #6693 : prevent memory leak during ConvertTable with a lot of partitions ## Empty changelog entries. The following PR had an empty `DESCRIPTION:` line. This generates an empty changelog line that needs to be removed manually. Please either provide a short entry, or remove `DESCRIPTION:` line completely. PR #6810 : Make CDC decoder an independent extension PR #6827 : Makefile changes to build CDC in builddir for pgoutput and wal2json. --------- Co-authored-by: Onur Tirtir <onurcantirtir@gmail.com>

) With this PR, we allow creating distributed tables with without specifying a shard key via create_distributed_table(). Here are the the important details about those tables: * Specifying `shard_count` is not allowed because it is assumed to be 1. * We mostly call such tables as "null shard-key" table in code / comments. * To avoid doing a breaking layout change in create_distributed_table(); instead of throwing an error, it will inform the user that `distribution_type` param is ignored unless it's explicitly set to NULL or 'h'. * `colocate_with` param allows colocating such null shard-key tables to each other. * We define this table type, i.e., NULL_SHARD_KEY_TABLE, as a subclass of DISTRIBUTED_TABLE because we mostly want to treat them as distributed tables in terms of SQL / DDL / operation support. * Metadata for such tables look like: - distribution method => DISTRIBUTE_BY_NONE - replication model => REPLICATION_MODEL_STREAMING - colocation id => **!=** INVALID_COLOCATION_ID (distinguishes from Citus local tables) * We assign colocation groups for such tables to different nodes in a round-robin fashion based on the modulo of "colocation id". Note that this PR doesn't care about DDL (except CREATE TABLE) / SQL / operation (i.e., Citus UDFs) support for such tables but adds a preliminary API.

onurctirtir commented Mar 3, 2023

View reviewed changes

onurctirtir requested review from onderkalaci and JelteF March 3, 2023 13:27

onurctirtir force-pushed the create-null-shard-key-tbl branch from ed967a6 to 2d88ba6 Compare March 3, 2023 14:18

onurctirtir requested a review from marcocitus March 7, 2023 09:53

onurctirtir force-pushed the create-null-shard-key-tbl branch 3 times, most recently from fef7e99 to 52d11cb Compare March 7, 2023 15:43

onurctirtir mentioned this pull request Mar 7, 2023

Refactor CreateDistributedTable() #6742

Merged

onurctirtir force-pushed the create-null-shard-key-tbl branch from 52d11cb to 6399baf Compare March 8, 2023 10:43

onurctirtir changed the base branch from main to clarify-dist-by-none March 8, 2023 10:44

onurctirtir force-pushed the clarify-dist-by-none branch from 15d8b48 to 538541d Compare March 8, 2023 23:29

onurctirtir force-pushed the create-null-shard-key-tbl branch 2 times, most recently from 8dce689 to b2d6fbf Compare March 9, 2023 09:17

onurctirtir force-pushed the clarify-dist-by-none branch 4 times, most recently from 7ee5b87 to bd2228a Compare March 9, 2023 11:53

marcocitus reviewed Mar 10, 2023

View reviewed changes

src/backend/distributed/commands/create_distributed_table.c Outdated Show resolved Hide resolved

marcocitus reviewed Mar 10, 2023

View reviewed changes

src/backend/distributed/utils/colocation_utils.c Outdated Show resolved Hide resolved

marcocitus reviewed Mar 10, 2023

View reviewed changes

Base automatically changed from clarify-dist-by-none to main March 10, 2023 10:55

onurctirtir force-pushed the create-null-shard-key-tbl branch 3 times, most recently from d778329 to efe2124 Compare March 13, 2023 09:13

onurctirtir commented Mar 13, 2023

View reviewed changes

Base automatically changed from decide-dist-params-in-create-citus-table to main March 14, 2023 11:24

onurctirtir force-pushed the create-null-shard-key-tbl branch from 3396fe0 to 4249163 Compare March 14, 2023 11:25

onderkalaci mentioned this pull request Mar 14, 2023

This implements MERGE phase-III #6696

Merged

onurctirtir force-pushed the create-null-shard-key-tbl branch 2 times, most recently from eb205d2 to 7b94697 Compare March 15, 2023 11:47

onurctirtir requested a review from marcocitus March 15, 2023 12:00

onurctirtir force-pushed the create-null-shard-key-tbl branch from 7b94697 to e16df14 Compare March 17, 2023 09:59

Add support for creating distributed tables with a null shard key

fbb8545

onurctirtir force-pushed the create-null-shard-key-tbl branch from e16df14 to fbb8545 Compare March 17, 2023 11:12

onurctirtir changed the base branch from main to null-shard-key-devel-protected March 17, 2023 11:15

onurctirtir merged commit a4cfc32 into null-shard-key-devel-protected Mar 17, 2023

onurctirtir deleted the create-null-shard-key-tbl branch March 17, 2023 11:17

hanefi mentioned this pull request Apr 14, 2023

Changelog entries for 11.3.0 #6856

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add support for creating distributed tables with a null shard key #6745

Add support for creating distributed tables with a null shard key #6745

onurctirtir commented Mar 3, 2023 •

edited

Loading

codecov bot commented Mar 3, 2023 •

edited

Loading

onurctirtir Mar 3, 2023 •

edited

Loading

marcocitus Mar 8, 2023

onurctirtir Mar 13, 2023 •

edited

Loading

marcocitus Mar 13, 2023

onurctirtir Mar 13, 2023 •

edited

Loading

marcocitus Mar 17, 2023

onurctirtir Mar 17, 2023

onurctirtir Mar 17, 2023

marcocitus Mar 17, 2023

marcocitus Mar 10, 2023

marcocitus Mar 10, 2023

marcocitus Mar 10, 2023

marcocitus left a comment

onurctirtir Mar 13, 2023

onurctirtir commented Mar 15, 2023 •

edited

Loading

onurctirtir commented Mar 15, 2023

marcocitus commented Mar 17, 2023

Add support for creating distributed tables with a null shard key #6745

Add support for creating distributed tables with a null shard key #6745

Conversation

onurctirtir commented Mar 3, 2023 • edited Loading

codecov bot commented Mar 3, 2023 • edited Loading

Codecov Report

onurctirtir Mar 3, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

onurctirtir Mar 13, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

onurctirtir Mar 13, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

marcocitus left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

onurctirtir commented Mar 15, 2023 • edited Loading

onurctirtir commented Mar 15, 2023

marcocitus commented Mar 17, 2023

onurctirtir commented Mar 3, 2023 •

edited

Loading

codecov bot commented Mar 3, 2023 •

edited

Loading

onurctirtir Mar 3, 2023 •

edited

Loading

onurctirtir Mar 13, 2023 •

edited

Loading

onurctirtir Mar 13, 2023 •

edited

Loading

onurctirtir commented Mar 15, 2023 •

edited

Loading