infoschema: optimize 'select count(*) from information_schema.tables' for v2 #55574

tiancaiamao · 2024-08-21T14:59:28Z

What problem does this PR solve?

Issue Number: close #55515
ref #50959

Problem Summary:

mysql> select count(*) from information_schema.tables;
ERROR 9006 (HY000): GC life time is shorter than transaction duration, transaction starts at 2024-08-20 07:13:02.343 +0800 CST, GC safe point is 2024-08-20 07:13:24.893 +0800 CST

What changed and how does it work?

Make a special optimization for reading INFORMATION_SCHEMA.TABLES, that is, if only table_name or table_schema are visited, make the query pure in-memory operation.

I tried to support select count(*) but found that the aggregation always use the first column for *, and in that case for INFORMATION_SCHEMA.TABLES, the column is table_catalog. And changing the select count(*) to using other columns seems more difficult than I thought.

Check List

Tests

Unit test
Integration test

Test with 1M tables:

mysql> select count(table_name) from tables;
+-------------------+
| count(table_name) |
+-------------------+
|           1007314 |
+-------------------+
1 row in set (1.54 sec)

Without this optimize:

mysql> select count(*) from tables;
+----------+
| count(*) |
+----------+
|  1007315 |
+----------+
1 row in set (9 min 50.39 sec)

Manual test (add detailed scripts or steps below)
No need to test
- I checked and no code files have been changed.

Side effects

Performance regression: Consumes more CPU
Performance regression: Consumes more Memory
Breaking backward compatibility

Documentation

Release note

Please refer to Release Notes Language Style Guide to write a quality release note.

None

…a.tables' for v2

tiprow · 2024-08-21T15:00:04Z

Hi @tiancaiamao. Thanks for your PR.

PRs from untrusted users cannot be marked as trusted with /ok-to-test in this repo meaning untrusted PR authors can never trigger tests themselves. Collaborators can still trigger tests on the PR using /test all.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

codecov · 2024-08-22T02:26:45Z

Codecov Report

Attention: Patch coverage is 93.68421% with 6 lines in your changes missing coverage. Please review.

Project coverage is 56.4415%. Comparing base (0583e84) to head (0c33bf2).
Report is 1 commits behind head on master.

Additional details and impacted files

@@                Coverage Diff                @@
##             master     #55574         +/-   ##
=================================================
- Coverage   72.7958%   56.4415%   -16.3543%     
=================================================
  Files          1588       1716        +128     
  Lines        443469     621216     +177747     
=================================================
+ Hits         322827     350624      +27797     
- Misses       100772     246943     +146171     
- Partials      19870      23649       +3779

Flag	Coverage Δ
integration	`37.2600% <84.2105%> (?)`
unit	`71.9841% <93.6842%> (+0.0776%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components	Coverage Δ
dumpling	`52.9567% <ø> (ø)`
parser	`∅ <ø> (∅)`
br	`51.8398% <ø> (+6.4086%)`	⬆️

tiancaiamao · 2024-08-22T04:47:42Z

/retest

tiprow · 2024-08-22T04:48:17Z

@tiancaiamao: Cannot trigger testing until a trusted user reviews the PR and leaves an /ok-to-test message.

In response to this:

/retest

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

tiancaiamao · 2024-08-23T00:24:21Z

/retest

tiprow · 2024-08-23T00:24:50Z

@tiancaiamao: Cannot trigger testing until a trusted user reviews the PR and leaves an /ok-to-test message.

In response to this:

/retest

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

tiancaiamao · 2024-08-23T01:29:09Z

/test check-dev

tiprow · 2024-08-23T01:29:30Z

@tiancaiamao: Cannot trigger testing until a trusted user reviews the PR and leaves an /ok-to-test message.

In response to this:

/test check-dev

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

tiancaiamao · 2024-08-23T02:09:01Z

/ok-to-test

bb7133 · 2024-08-23T20:21:30Z

Is select count(*) from information_schema.tables; widely used? Just curious.

lance6716 · 2024-08-26T01:46:48Z

pkg/executor/infoschema_reader.go

+			IterateAllTableItems(visit func(infoschema.TableItem) bool)
+		})
+		if ok {
+			if x := ctx.Value("cover-check"); x != nil {


maybe use a failpoint

IMO, failpoint is not as convient sometimes.

It works globally, so there's no good way for precise control with a specific query.

It requires make failpoint-enable, an extra step to enable the test

It requires make failpoint-enable, an extra step to enable the test

I have added this feature for failpoint, it can be configurated with our other testing command line arguments like --tags=intest or -race https://github.com/pingcap/failpoint?tab=readme-ov-file#quick-start-use-failpoint-toolexec

tangenta · 2024-08-27T14:12:00Z

tests/integrationtest/t/executor/infoschema_reader.test

+select table_name from information_schema.tables where table_name = 't';
+explain format='brief' select table_name, table_schema from information_schema.tables;
+select count(*) from information_schema.tables where table_name = 't';
+select count(table_name) from information_schema.tables where table_name = 't';


Suggested change

select count(table_name) from information_schema.tables where table_name = 't';

select count(table_name) from information_schema.tables where table_name = 't';

ti-chi-bot · 2024-08-27T14:12:26Z

[LGTM Timeline notifier]

Timeline:

2024-08-26 01:49:11.122248659 +0000 UTC m=+748546.256698769: ☑️ agreed by lance6716.
2024-08-27 14:12:25.522595228 +0000 UTC m=+879540.657045346: ☑️ agreed by tangenta.

fixdb

+1

ti-chi-bot · 2024-09-03T01:02:48Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: fixdb, lance6716, tangenta

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

~~OWNERS~~ [fixdb,lance6716,tangenta]
~~pkg/infoschema/OWNERS~~ [tangenta]
~~pkg/planner/OWNERS~~ [fixdb]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

infoschema: optimize 'select count(table_name) from information_schem…

080637d

…a.tables' for v2

ti-chi-bot bot added do-not-merge/needs-linked-issue release-note-none Denotes a PR that doesn't merit a release note. size/L Denotes a PR that changes 100-499 lines, ignoring generated files. labels Aug 21, 2024

ti-chi-bot bot added do-not-merge/needs-triage-completed and removed do-not-merge/needs-linked-issue labels Aug 22, 2024

tiancaiamao added 2 commits August 22, 2024 09:38

Merge branch 'master' into optimize

2aebbf1

merge master

f103080

ti-chi-bot bot added sig/planner SIG: Planner and removed do-not-merge/needs-triage-completed labels Aug 22, 2024

add privilege check

db493c8

address comment

fd3cf1e

tiancaiamao changed the title ~~infoschema: optimize 'select count(table_name) from information_schema.tables' for v2~~ infoschema: optimize 'select count(*) from information_schema.tables' for v2 Aug 22, 2024

tiancaiamao added 2 commits August 22, 2024 19:16

update comment

26a162c

address comment

60a52f0

ti-chi-bot bot added the ok-to-test Indicates a PR is ready to be tested. label Aug 23, 2024

tiancaiamao requested review from lance6716 and tangenta August 23, 2024 02:52

lance6716 approved these changes Aug 26, 2024

View reviewed changes

ti-chi-bot bot added the needs-1-more-lgtm Indicates a PR needs 1 more LGTM. label Aug 26, 2024

tangenta reviewed Aug 27, 2024

View reviewed changes

tangenta approved these changes Aug 27, 2024

View reviewed changes

ti-chi-bot bot added lgtm and removed needs-1-more-lgtm Indicates a PR needs 1 more LGTM. labels Aug 27, 2024

fixdb approved these changes Sep 3, 2024

View reviewed changes

ti-chi-bot bot added the approved label Sep 3, 2024

tiancaiamao added 2 commits September 3, 2024 15:24

Merge branch 'master' into optimize

83494b0

Merge master

0c33bf2

ti-chi-bot bot merged commit dd180de into pingcap:master Sep 3, 2024
23 checks passed

tiancaiamao deleted the optimize branch September 3, 2024 08:33

This was referenced Sep 3, 2024

tidb/pull/55574 does not handle snapshot infoschema version #55827

Closed

infoschema: fix snapshot read bug for information_schema.tables #55830

Merged

Flaky test TestShowViewPriv #55867

Closed

tangenta mentioned this pull request Sep 27, 2024

planner: make pattern match case-insensitive for some infoschema tables #56378

Merged

13 tasks

joechenrh mentioned this pull request Nov 19, 2024

planner: fix count(*) return wrong value for information_schema.tables #57506

Merged

13 tasks

ti-chi-bot mentioned this pull request Nov 20, 2024

planner: fix count(*) return wrong value for information_schema.tables (#57506) #57549

Merged

13 tasks

tiancaiamao mentioned this pull request Nov 27, 2024

executor: fix query infoschema.tables table_schema/table_name with filter condition #57746

Merged

13 tasks

ti-chi-bot mentioned this pull request Nov 27, 2024

executor: fix query infoschema.tables table_schema/table_name with filter condition (#57746) #57771

Merged

13 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

infoschema: optimize 'select count(*) from information_schema.tables' for v2 #55574

infoschema: optimize 'select count(*) from information_schema.tables' for v2 #55574

tiancaiamao commented Aug 21, 2024 •

edited

Loading

tiprow bot commented Aug 21, 2024

codecov bot commented Aug 22, 2024 •

edited

Loading

tiancaiamao commented Aug 22, 2024

tiprow bot commented Aug 22, 2024

tiancaiamao commented Aug 23, 2024

tiprow bot commented Aug 23, 2024

tiancaiamao commented Aug 23, 2024

tiprow bot commented Aug 23, 2024

tiancaiamao commented Aug 23, 2024

bb7133 commented Aug 23, 2024

lance6716 Aug 26, 2024

tiancaiamao Aug 27, 2024

lance6716 Aug 27, 2024

tangenta Aug 27, 2024

ti-chi-bot bot commented Aug 27, 2024

fixdb left a comment

ti-chi-bot bot commented Sep 3, 2024

	select count(table_name) from information_schema.tables where table_name = 't';
	select count(table_name) from information_schema.tables where table_name = 't';

infoschema: optimize 'select count(*) from information_schema.tables' for v2 #55574

infoschema: optimize 'select count(*) from information_schema.tables' for v2 #55574

Conversation

tiancaiamao commented Aug 21, 2024 • edited Loading

What problem does this PR solve?

What changed and how does it work?

Check List

Release note

tiprow bot commented Aug 21, 2024

codecov bot commented Aug 22, 2024 • edited Loading

Codecov Report

tiancaiamao commented Aug 22, 2024

tiprow bot commented Aug 22, 2024

tiancaiamao commented Aug 23, 2024

tiprow bot commented Aug 23, 2024

tiancaiamao commented Aug 23, 2024

tiprow bot commented Aug 23, 2024

tiancaiamao commented Aug 23, 2024

bb7133 commented Aug 23, 2024

lance6716 Aug 26, 2024

Choose a reason for hiding this comment

tiancaiamao Aug 27, 2024

Choose a reason for hiding this comment

lance6716 Aug 27, 2024

Choose a reason for hiding this comment

tangenta Aug 27, 2024

Choose a reason for hiding this comment

ti-chi-bot bot commented Aug 27, 2024

[LGTM Timeline notifier]

fixdb left a comment

Choose a reason for hiding this comment

ti-chi-bot bot commented Sep 3, 2024

tiancaiamao commented Aug 21, 2024 •

edited

Loading

codecov bot commented Aug 22, 2024 •

edited

Loading