Fixing join tests with analyzer by vdimir · Pull Request #49555 · ClickHouse/ClickHouse

vdimir · 2023-05-05T13:06:38Z

Changelog category (leave one):

Not for changelog (changelog entry is not required)

TODO:

Fix join_use_nulls in analyzer #49359

Fixed reference files:

01353_low_cardinality_join_types

 -- Without Analyzer LowCardinality(String)
 -- With Analyzer: String
 SELECT toTypeName(s) FROM (SELECT toLowCardinality('a') as s) t1 JOIN (SELECT 'a' as s) t2 USING (s);
 
 -- Without Analyzer LowCardinality(String)
 -- With Analyzer: String
 SELECT toTypeName(t1.s) FROM (SELECT toLowCardinality('a') as s) t1 JOIN (SELECT 'a' as s) t2 USING (s);

 -- Without Analyzer String
 -- With Analyzer: String
 SELECT toTypeName(t2.s) FROM (SELECT toLowCardinality('a') as s) t1 JOIN (SELECT 'a' as s) t2 USING (s);

02267_join_dup_columns_issue36199

-- Without Analyzer return only columns `count(), x` from right table
-- With Analyzer return only columns `count('x'), count('y'), x` from right table
SELECT * FROM ( SELECT 2 AS x ) AS t1 RIGHT JOIN ( SELECT count('x'), count('y'), 2 AS x ) AS t2 ON t1.x = t2.x;

02242_join_rocksdb

Fixed bug and adjusted error code in test
01049_join_low_card_bug_long

Fix bug with query:
```
SELECT t.t from (SELECT 1 as t) as a FULL JOIN (SELECT 1 as t) as t USING t;
```
adjust reference file (columns after USING have type of supertype of the source columns), also it needs fix from Fix join_use_nulls in analyzer #49359 (so, it's not removed from broken_tests.txt)

01721_join_implicit_cast_long
Adjust reference.

Not sure about the change:

 -- Before: Nullable
 -- With analyzer: not-Nullable
 SELECT toTypeName(a) FROM (SELECT 1 as a ) t1 RIGHT JOIN  (SELECT 1 as a) t2 USING (a) SETTINGS join_use_nulls = 1;

01062_pm_all_join_with_block_continuation
~~Fixed test file, JOIN using alias from external scope.~~ Removed this fix from the PR, will in another one.
02000_join_on_const
Fixed test file, ... ANY JOIN ... ON 1 == 1 is forbidden
02382_join_and_filtering_set
Copy missing corresponding code from Interpreter, adjust test (removed with cube)
01890_materialized_distributed_join
Function getHeaderForProcessingStage handle join on top of query tree

robot-clickhouse-ci-2 · 2023-05-05T13:33:03Z

This is an automated comment for commit 3aac663 with description of existing statuses. It's updated for the latest CI running

❌ Click here to open a full report in a separate page

Successful checks

Check name	Description	Status
AST fuzzer	Runs randomly generated queries to catch program errors. The build type is optionally given in parenthesis. If it fails, ask a maintainer for help	✅ success
CI running	A meta-check that indicates the running CI. Normally, it's in success or pending state. The failed status indicates some problems with the PR	✅ success
ClickHouse build check	Builds ClickHouse in various configurations for use in further steps. You have to fix the builds that fail. Build logs often has enough information to fix the error, but you might have to reproduce the failure locally. The cmake options can be found in the build log, grepping for cmake. Use these options and follow the general build process	✅ success
Compatibility check	Checks that clickhouse binary runs on distributions with old libc versions. If it fails, ask a maintainer for help	✅ success
Docker image for servers	The check to build and optionally push the mentioned image to docker hub	✅ success
Fast test	Normally this is the first check that is ran for a PR. It builds ClickHouse and runs most of stateless functional tests, omitting some. If it fails, further checks are not started until it is fixed. Look at the report to see which tests fail, then reproduce the failure locally as described here	✅ success
Flaky tests	Checks if new added or modified tests are flaky by running them repeatedly, in parallel, with more randomization. Functional tests are run 100 times with address sanitizer, and additional randomization of thread scheduling. Integrational tests are run up to 10 times. If at least once a new test has failed, or was too long, this check will be red. We don't allow flaky tests, read the doc	✅ success
Install packages	Checks that the built packages are installable in a clear environment	✅ success
Mergeable Check	Checks if all other necessary checks are successful	✅ success
Performance Comparison	Measure changes in query performance. The performance test report is described in detail here. In square brackets are the optional part/total tests	✅ success
Push to Dockerhub	The check for building and pushing the CI related docker images to docker hub	✅ success
SQLTest	There's no description for the check yet, please add it to tests/ci/ci_config.py:CHECK_DESCRIPTIONS	✅ success
SQLancer	Fuzzing tests that detect logical bugs with SQLancer tool	✅ success
Sqllogic	Run clickhouse on the sqllogic test set against sqlite and checks that all statements are passed	✅ success
Stateful tests	Runs stateful functional tests for ClickHouse binaries built in various configurations -- release, debug, with sanitizers, etc	✅ success
Stateless tests	Runs stateless functional tests for ClickHouse binaries built in various configurations -- release, debug, with sanitizers, etc	✅ success
Stress test	Runs stateless functional tests concurrently from several clients to detect concurrency-related errors	✅ success
Style Check	Runs a set of checks to keep the code style clean. If some of tests failed, see the related log from the report	✅ success
Unit tests	Runs the unit tests for different release types	✅ success

Check name	Description	Status
Integration tests	The integration tests report. In parenthesis the package type is given, and in square brackets are the optional part/total tests	❌ failure
Upgrade check	Runs stress tests on server version from last release and then tries to upgrade it to the version from the PR. It checks if the new server can successfully startup without any errors, crashes or sanitizer asserts	❌ failure

vdimir · 2023-05-10T11:25:43Z

Tests failed because #49359 is also required, I'll temporarily add commits from that branch here.

vdimir · 2023-05-12T12:20:02Z

https://s3.amazonaws.com/clickhouse-test-reports/49555/1e50db004d10d0322494635f60897e7c1a963c84/fuzzer_astfuzzerasan/report.html

vdimir · 2023-08-03T15:04:22Z

Stateless tests flaky check (asan) — fail: 21, passed: 193 Details
01062_pm_all_join_with_block_continuation: [ FAIL ] 147.69 sec. - Test runs too long (> 60s). Make it faster

Normally it takes <50sec:
https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBxdWFudGlsZXMoMC41LCAwLjksIDAuOTUpKHRlc3RfZHVyYXRpb25fbXMgLyAxMDAwKSBGUk9NIGNoZWNrcwpXSEVSRSB0ZXN0X25hbWUgPSAnMDEwNjJfcG1fYWxsX2pvaW5fd2l0aF9ibG9ja19jb250aW51YXRpb24nCkFORCBjaGVja19uYW1lIGxpa2UgJyVhc2FuJScKQU5EIGNoZWNrX3N0YXJ0X3RpbWUgPD0gbm93KCkgLSBJTlRFUlZBTCAxMCBEQVkKQU5EIHB1bGxfcmVxdWVzdF9udW1iZXIgPSAwCgo=

Locally new and old versions of tests are approximately the same.

Temporarily reverted changes in this test 01062_pm_all_join_with_block_continuation

vdimir · 2023-08-04T09:40:20Z

Stateless tests flaky check (asan) — fail: 1, passed: 260 Details

01721_join_implicit_cast_long: [ FAIL ] 601.03 sec. - Timeout!

It doesn't have long tag in the code, I'm not sure if we still count test as long by filename.
It became twice longer that it is (added run it with and without analyzer). Another option is just to split it into two tests.

All in all, this PR is finished it contains fixes for several tests. What was done for each test is described in the top-level PR comment, and also each one is addressed in separate commit.

@novikd you may take a look

alexey-milovidov · 2023-08-12T00:29:17Z

The new test appears to be longer than 600 seconds under some settings.

novikd

In general LGTM

novikd · 2023-09-29T17:53:34Z

src/Interpreters/getHeaderForProcessingStage.cpp

-                removeJoin(*query->as<ASTSelectQuery>(), new_rewriter_result, context);
+                {
+                    if (!query_info.planner_context)
+                        throw Exception(ErrorCodes::LOGICAL_ERROR, "Query is not analyzed");


Maybe it's possible to make a more meaningful log message.

Updated to Query is not analyzed: no planner context. Either way, it's LOGICAL_ERROR and it doesn't matter a lot for a user, but rather for developer.

novikd · 2023-09-29T17:55:06Z

src/Planner/PlannerJoinTree.cpp

+            /// Note: it's also doesn't work with the read-in-order optimization.
+            /// No checks here because read in order is not applied if we have `CreateSetAndFilterOnTheFlyStep` in the pipeline between the reading and sorting steps.
+            bool has_non_const_keys = has_non_const(left_plan.getCurrentDataStream().header, join_clause.key_names_left)
+                && has_non_const(right_plan.getCurrentDataStream().header, join_clause.key_names_right);


Maybe || instead of &&?

Both sides should have non constant keys, otherwise SortingTransform on corresponding side emit rows before whole table had been read and it breaks CreateSetAndFilterOnTheFlyStep

src/Storages/HDFS/StorageHDFSCluster.cpp

src/Storages/StorageS3Cluster.cpp

novikd · 2023-09-29T17:57:17Z

tests/queries/0_stateless/01049_join_low_card_bug_long.sql.j2


 SELECT lc, toTypeName(lc)  FROM l_lc AS l RIGHT JOIN r_lc AS r USING (x) ORDER BY l.lc;

-SELECT lowCardinalityKeys(lc.lc) FROM r FULL JOIN l_lc as lc USING (lc) ORDER BY lowCardinalityKeys(lc.lc);


Why is it removed?

In new analyzer lc.lc became String (because r contains String and we take supertype as a result type), so lowCardinalityKeys is not applicable. We can leave this query only for old analyzer, but I don't think it's essential.

vdimir · 2023-10-10T09:13:42Z

Stress test (tsan) — check_status.tsv doesn't exists Details
Upgrade check (tsan) — check_status.tsv doesn't exists Details

#49076

Stateless tests (release, analyzer) — fail: 1, passed: 5649, skipped: 48 Details

02886_missed_json_subcolumns probably related, will check

UPD: this check was after checing can_be_not_found after merge, fixed

ClickHouse/src/Analyzer/Passes/QueryAnalysisPass.cpp

Lines 2426 to 2441 in a799d4e

    
           if (auto * column = compound_expression->as<ColumnNode>()) 
        
           { 
        
               const DataTypePtr & column_type = column->getColumn().getTypeInStorage(); 
        
               if (column_type->getTypeId() == TypeIndex::Object) 
        
               { 
        
                   const auto * object_type = checkAndGetDataType<DataTypeObject>(column_type.get()); 
        
                   if (object_type->getSchemaFormat() == "json" && object_type->hasNullableSubcolumns()) 
        
                   { 
        
                       QueryTreeNodePtr constant_node_null = std::make_shared<ConstantNode>(Field()); 
        
                       return constant_node_null; 
        
                   } 
        
               } 
        
           } 
        
           if (can_be_not_found) 
        
               return {};

Fixed test file, JOIN using alias from external scope

Copy missing corresponding code from Interpreter, adjust test (removed with cube)

Function getHeaderForProcessingStage handle join on top of query tree

This reverts commit 88865ac.

vdimir · 2023-10-13T10:29:24Z

Integration tests (release) [3/4] — fail: 1, passed: 603 Details

Seems test_replicated_zero_copy_projection_mutation/test.py::test_hardlinks_preserved_when_projection_dropped is a bit flaky.
No other failures (ci db), but issue related to timeouts in tests

Details

We deleted objects after timeout:

2023-10-12 12:23:59 [ 601 ] DEBUG : http://172.16.11.8:9001 "GET /root?delimiter=&encoding-type=url&list-type=2&max-keys=1000&prefix=data%2F HTTP/1.1" 200 0 (connectionpool.py:546, _make_request)
2023-10-12 12:23:59 [ 601 ] INFO : list_objects (20): ['data/bvw/wrljevlduzmgtpddbnjdicmjbqmul', 'data/gar/bdzxvruhcbojwnssexrlvsuoknmyg', 'data/ixn/uqsdxqevkpebrpidxuywczmsmwhlb', 'data/kol/jzwsrnushtokftdzaloolthtxldvp', 'data/kvs/uhzthybwkhsxuiteqoyvdlniitnfb', 'data/lhi/ctdysdzbymnllpxcqffbyrnxjonwf', 'data/loi/nkwrplbttcgzamiwqmedqbfgqcxvk', 'data/lsx/iyuqumlazclzrdsitaobgcbdzookw', 'data/mbx/gqvwlivusmkrehndgnbnjkochywdc', 'data/mqk/dljgyadfnccqijfgudkkgnkssrnej', 'data/nzb/fhfkriolhlgrjcskicmcyivmfjegy', 'data/ood/deqohcrddwnsademgifqbmvunctgq', 'data/pnd/kbdarptmoapjfglrklaeevnaqtkex', 'data/rey/tesbqzzygjsnwrvvrbvesjlscbetv', 'data/seq/wlsptkdxjzjxwnvesxhuyltndvbcp', 'data/sya/uaggjvudvwbtgvbvqmromayzzupdc', 'data/vie/vtakpllynnxwkykfmsghonyqpeqra', 'data/vla/bsrahqodprsluqzngqkuivtmrcebm', 'data/vtd/vehexvssfhqseohjruybicspupwcl', 'data/xts/jusayraucumkmwegngwsgdyenokeu'] (test.py:75, list_objects)
2023-10-12 12:24:00 [ 601 ] DEBUG : http://172.16.11.8:9001 "GET /root?delimiter=&encoding-type=url&list-type=2&max-keys=1000&prefix=data%2F HTTP/1.1" 200 0 (connectionpool.py:546, _make_request)
2023-10-12 12:24:00 [ 601 ] INFO : list_objects (20): ['data/bvw/wrljevlduzmgtpddbnjdicmjbqmul', 'data/gar/bdzxvruhcbojwnssexrlvsuoknmyg', 'data/ixn/uqsdxqevkpebrpidxuywczmsmwhlb', 'data/kol/jzwsrnushtokftdzaloolthtxldvp', 'data/kvs/uhzthybwkhsxuiteqoyvdlniitnfb', 'data/lhi/ctdysdzbymnllpxcqffbyrnxjonwf', 'data/loi/nkwrplbttcgzamiwqmedqbfgqcxvk', 'data/lsx/iyuqumlazclzrdsitaobgcbdzookw', 'data/mbx/gqvwlivusmkrehndgnbnjkochywdc', 'data/mqk/dljgyadfnccqijfgudkkgnkssrnej', 'data/nzb/fhfkriolhlgrjcskicmcyivmfjegy', 'data/ood/deqohcrddwnsademgifqbmvunctgq', 'data/pnd/kbdarptmoapjfglrklaeevnaqtkex', 'data/rey/tesbqzzygjsnwrvvrbvesjlscbetv', 'data/seq/wlsptkdxjzjxwnvesxhuyltndvbcp', 'data/sya/uaggjvudvwbtgvbvqmromayzzupdc', 'data/vie/vtakpllynnxwkykfmsghonyqpeqra', 'data/vla/bsrahqodprsluqzngqkuivtmrcebm', 'data/vtd/vehexvssfhqseohjruybicspupwcl', 'data/xts/jusayraucumkmwegngwsgdyenokeu'] (test.py:75, list_objects)
2023-10-12 12:24:00 [ 601 ] DEBUG : Executing query DROP TABLE IF EXISTS test_hardlinks_preserved_when_projection_dropped SYNC on node1 (cluster.py:3409, query)
2023-10-12 12:24:00 [ 601 ] DEBUG : Executing query DROP TABLE IF EXISTS test_hardlinks_preserved_when_projection_dropped SYNC on node2 (cluster.py:3409, query)

2023.10.12 12:24:00.270615 [ 371 ] {} <Debug> S3ObjectStorage: Objects with paths [data/vie/vtakpllynnxwkykfmsghonyqpeqra, data/nzb/fhfkriolhlgrjcskicmcyivmfjegy, data/rey/tesbqzzygjsnwrvvrbvesjlscbetv, data/lhi/ctdysdzbymnllpxcqffbyrnxjonwf, data/bvw/wrljevlduzmgtpddbnjdicmjbqmul, data/sya/uaggjvudvwbtgvbvqmromayzzupdc, data/seq/wlsptkdxjzjxwnvesxhuyltndvbcp, data/vtd/vehexvssfhqseohjruybicspupwcl, data/mqk/dljgyadfnccqijfgudkkgnkssrnej] were removed from S3

And expected objects:

list_objects (2): ['data/gar/bdzxvruhcbojwnssexrlvsuoknmyg', 'data/mbx/gqvwlivusmkrehndgnbnjkochywdc'] (test.py:75, list_objects)

Was deleted only after test was finished and table was dropped.

node1/logs/clickhouse-server.log
517:2023.10.12 12:22:53.716436 [ 12 ] {ea2e0500-08fa-4c29-95f7-4297baca346a} <Debug> executeQuery: (from 172.16.11.1:55232) DROP TABLE IF EXISTS test_hardlinks_preserved_when_projection_dropped SYNC (stage: Complete)
2037:2023.10.12 12:24:00.196979 [ 12 ] {8db3cd4d-2ed7-4a21-9e47-2ae7feff10df} <Debug> executeQuery: (from 172.16.11.1:55710) DROP TABLE IF EXISTS test_hardlinks_preserved_when_projection_dropped SYNC (stage: Complete)

node2/logs/clickhouse-server.log
907:2023.10.12 12:24:00.260354 [ 361 ] {205de67e-686c-4b76-8d04-21e08a19b434} <Debug> executeQuery: (from 172.16.11.1:64916) DROP TABLE IF EXISTS test_hardlinks_preserved_when_projection_dropped SYNC (stage: Complete)
...

2023.10.12 12:24:00.201130 [ 163 ] {} <Debug> S3ObjectStorage: Objects with paths [data/mbx/gqvwlivusmkrehndgnbnjkochywdc] were removed from S3
...
2023.10.12 12:24:00.275343 [ 371 ] {} <Debug> S3ObjectStorage: Objects with paths [data/gar/bdzxvruhcbojwnssexrlvsuoknmyg] were removed from S3

So I believe no bug here

vdimir added the do not test disable testing on pull request label May 5, 2023

robot-ch-test-poll4 added the pr-not-for-changelog This PR should not be mentioned in the changelog label May 5, 2023

vdimir mentioned this pull request May 5, 2023

[Umbrella] Analyzer, Planner migration #42648

Closed

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch from 8a9ca47 to 8e92730 Compare May 8, 2023 15:26

vdimir removed the do not test disable testing on pull request label May 8, 2023

novikd self-assigned this May 9, 2023

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch 2 times, most recently from b596514 to 9a975bf Compare May 9, 2023 16:41

vdimir marked this pull request as draft May 10, 2023 11:25

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch from 9a975bf to 1e50db0 Compare May 10, 2023 15:58

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch 2 times, most recently from 97115ca to 7c80861 Compare June 1, 2023 10:49

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch from 18c485c to 4af8d68 Compare August 2, 2023 10:05

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch from 4af8d68 to 964c7f9 Compare August 3, 2023 15:24

vdimir marked this pull request as ready for review August 4, 2023 09:40

vdimir requested a review from novikd August 4, 2023 09:47

novikd reviewed Sep 29, 2023

View reviewed changes

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch from 964c7f9 to 4cfe600 Compare October 5, 2023 12:42

vdimir added 5 commits October 11, 2023 10:54

Fix 02267_join_dup_columns_issue36199

57dfe29

Fix 01353_low_cardinality_join_types

7a9c3ca

Fix 02242_join_rocksdb

1c70e04

[wip] Fixing 01049_join_low_card_bug_long

2875ea4

Fix 01049_join_low_card_bug_long

a799d4e

vdimir added 8 commits October 12, 2023 10:45

Fix 01721_join_implicit_cast_long

624f386

Fix 01062_pm_all_join_with_block_continuation

fcb116a

Fixed test file, JOIN using alias from external scope

Fix 02000_join_on_const

16eeec6

Fix 02382_join_and_filtering_set

84e838d

Copy missing corresponding code from Interpreter, adjust test (removed with cube)

Fix 01890_materialized_distributed_join

94fa777

Function getHeaderForProcessingStage handle join on top of query tree

fix style check

dcdcdc5

Revert "Fix 01062_pm_all_join_with_block_continuation"

ef989fd

This reverts commit 88865ac.

fix

3aac663

vdimir force-pushed the vdimir/analyzer_fix_join_tests branch from 4cfe600 to 3aac663 Compare October 12, 2023 10:47

vdimir merged commit e64abfc into master Oct 13, 2023

vdimir deleted the vdimir/analyzer_fix_join_tests branch October 13, 2023 10:30


		SELECT lc, toTypeName(lc) FROM l_lc AS l RIGHT JOIN r_lc AS r USING (x) ORDER BY l.lc;

		SELECT lowCardinalityKeys(lc.lc) FROM r FULL JOIN l_lc as lc USING (lc) ORDER BY lowCardinalityKeys(lc.lc);

Conversation

vdimir commented May 5, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Changelog category (leave one):

Uh oh!

robot-clickhouse-ci-2 commented May 5, 2023 • edited by robot-ch-test-poll2 Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vdimir commented May 10, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vdimir commented May 12, 2023

Uh oh!

vdimir commented Aug 3, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vdimir commented Aug 4, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

alexey-milovidov commented Aug 12, 2023

Uh oh!

novikd left a comment

Choose a reason for hiding this comment

Uh oh!

novikd Sep 29, 2023

Choose a reason for hiding this comment

Uh oh!

vdimir Oct 5, 2023

Choose a reason for hiding this comment

Uh oh!

novikd Sep 29, 2023

Choose a reason for hiding this comment

Uh oh!

vdimir Oct 5, 2023

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

novikd Sep 29, 2023

Choose a reason for hiding this comment

Uh oh!

vdimir Oct 5, 2023

Choose a reason for hiding this comment

Uh oh!

vdimir commented Oct 10, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vdimir commented Oct 13, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

vdimir commented May 5, 2023 •

edited

Loading

robot-clickhouse-ci-2 commented May 5, 2023 •

edited by robot-ch-test-poll2

Loading

vdimir commented May 10, 2023 •

edited

Loading

vdimir commented Aug 3, 2023 •

edited

Loading

vdimir commented Aug 4, 2023 •

edited

Loading

vdimir commented Oct 10, 2023 •

edited

Loading