Skip to content

Conversation

terry1purcell
Copy link
Contributor

@terry1purcell terry1purcell commented Oct 29, 2024

What problem does this PR solve?

Issue Number: close #55632

Problem Summary:

What changed and how does it work?

MySQL will increase the cost of a table scan if FORCE INDEX is used in the query. TiDB does not differentiate between USE and FORCE in the query. This enhancement will add a cost penalty to a full table scan if USE or FORCE is used in the query. It will also apply to ANY full table scan in the query, and not only to the table that the USE/FORCE is applied to. This is in response to customer issues where USE/FORCE is specified, but they still see table scans chosen within their query.

If a USE/FORCE is specified against the PRIMARY index, and there are no other predicates (such that the scan is a full table scan of that primary) - this enhancement will increase that table scan cost.

The cost penalty does not apply to a table range scan.

Check List

Tests

  • Unit test
  • Integration test
  • 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

  • Affects user behaviors
  • Contains syntax changes
  • Contains variable changes
  • Contains experimental features
  • Changes MySQL compatibility

Release note

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

None

@ti-chi-bot ti-chi-bot bot added release-note-none Denotes a PR that doesn't merit a release note. size/M Denotes a PR that changes 30-99 lines, ignoring generated files. sig/planner SIG: Planner labels Oct 29, 2024
Copy link

tiprow bot commented Oct 29, 2024

Hi @terry1purcell. 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.

@hawkingrei hawkingrei self-requested a review October 29, 2024 02:59
@hawkingrei
Copy link
Member

/ok-to-test

@ti-chi-bot ti-chi-bot bot added the ok-to-test Indicates a PR is ready to be tested. label Oct 29, 2024
Copy link

codecov bot commented Oct 29, 2024

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 74.8237%. Comparing base (9812d85) to head (d8c4bd8).
Report is 128 commits behind head on master.

Additional details and impacted files
@@               Coverage Diff                @@
##             master     #56928        +/-   ##
================================================
+ Coverage   73.1839%   74.8237%   +1.6398%     
================================================
  Files          1671       1720        +49     
  Lines        460731     476167     +15436     
================================================
+ Hits         337181     356286     +19105     
+ Misses       102827      97606      -5221     
- Partials      20723      22275      +1552     
Flag Coverage Δ
integration 49.5419% <100.0000%> (?)
unit 72.2762% <100.0000%> (-0.0290%) ⬇️

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

Components Coverage Δ
dumpling 52.6910% <ø> (ø)
parser ∅ <ø> (∅)
br 61.6650% <ø> (+15.6606%) ⬆️

@terry1purcell terry1purcell changed the title planner: Use/force to trigger skyline pruning planner: Use/force to apply prefer range scan Oct 29, 2024
@hawkingrei
Copy link
Member

/retest

@ti-chi-bot ti-chi-bot bot added size/L Denotes a PR that changes 100-499 lines, ignoring generated files. and removed size/M Denotes a PR that changes 30-99 lines, ignoring generated files. labels Nov 1, 2024
@ti-chi-bot ti-chi-bot bot added size/XL Denotes a PR that changes 500-999 lines, ignoring generated files. and removed size/L Denotes a PR that changes 100-499 lines, ignoring generated files. labels Nov 2, 2024
Copy link

ti-chi-bot bot commented Dec 5, 2024

@qw4990: Your lgtm message is repeated, so it is ignored.

In response to this:

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.

@terry1purcell
Copy link
Contributor Author

/retest

Copy link
Contributor

@AilinKid AilinKid left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Rest LGTM

Copy link

ti-chi-bot bot commented Dec 9, 2024

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: AilinKid, qw4990

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:

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

@ti-chi-bot ti-chi-bot bot added lgtm and removed needs-1-more-lgtm Indicates a PR needs 1 more LGTM. labels Dec 9, 2024
Copy link

ti-chi-bot bot commented Dec 9, 2024

[LGTM Timeline notifier]

Timeline:

  • 2024-11-25 07:17:15.077214159 +0000 UTC m=+448022.696868672: ☑️ agreed by qw4990.
  • 2024-12-09 08:16:52.135339716 +0000 UTC m=+253602.224142258: ☑️ agreed by AilinKid.

@ti-chi-bot ti-chi-bot bot merged commit cdfdb8a into pingcap:master Dec 9, 2024
23 of 24 checks passed
@terry1purcell terry1purcell deleted the forced branch December 10, 2024 18:48
@terry1purcell terry1purcell restored the forced branch December 20, 2024 15:56
@terry1purcell terry1purcell added the needs-cherry-pick-release-8.5 Should cherry pick this PR to release-8.5 branch. label Dec 20, 2024
@ti-chi-bot
Copy link
Member

In response to a cherrypick label: new pull request created to branch release-8.5: #58444.

ti-chi-bot pushed a commit to ti-chi-bot/tidb that referenced this pull request Dec 20, 2024
crazycs520 pushed a commit to crazycs520/tidb that referenced this pull request Jun 4, 2025
…_pr=95

* executor: fix a bug that global temporary table send cop request (pingcap#588…
* statistics: fix the panic when to async load stats with dropped index …
* executor: fix prepared protocol charset (pingcap#58872) (pingcap#58931)
* *: Update client-go and verify all read ts (pingcap#58909)
* integration test: fix test case "br_pitr" (pingcap#58876)
* session: add indexes for `mysql.analyze_jobs` (pingcap#58134) (pingcap#58355)
* ddl: fix args count for modify column (pingcap#58855) (pingcap#58858)
* planner: correct plan when scan tidb related cluster table with KeepOr…
* planner: Fix vector not truncated after CBO (pingcap#58809) (pingcap#58844)
* ddl: Fix vector index for high dimensional vectors (pingcap#58717) (pingcap#58835)
* ddl: Fix issue with concurrent update getting reverted by BackfillData…
* statistics: stats cache set default quota as 20%  (pingcap#58013) (pingcap#58817)
* executor: change the evaluation order of columns in `Update` and `Inse…
* statistics: add recover to protect background task (pingcap#58739) (pingcap#58767)
* ttl: fix the infinite waiting for delRateLimiter when `tidb_ttl_delete…
* ttl: reduce some warnings logs when locking TTL tasks (pingcap#58306) (pingcap#58783)
* ttl: retry the rows when del rate limiter returns error in delWorker (…
* ttl: reschedule task to other instances when shrinking worker (pingcap#57703) (pingcap#58778)
* ttl: fix the issue that one task losing heartbeat will block other tas…
* ttl: fix the issue that the task is not cancelled after transfering ow…
* ddl: fix job state overridden when concurrent updates don't overlap in…
* ttl: set the job history status to `cancelled` if it's removed in GC a…
* ttl: fix the timezone issue and panic in the caller of `getSession` (#…
* ddl: fix version syncer doesn't print who hasn't synced on partial syn…
* ttl: fix the issue that `DROP TABLE` / `ALTER TABLE` will keep job run…
* br/stream: allow pitr to create oversized indices (pingcap#58433) (pingcap#58527)
* ttl: set a result for timeout scan task during shrinking scan worker (…
* executor: fix time zone issue when querying slow log (pingcap#58455) (pingcap#58577)
* table: fix the issue that the default value for `BIT` column is wrong …
* statistics: temporarily skip handling errors for DDL events (pingcap#58609) (pingcap#58634)
* sessionctx: fix null max value to leading wrong warning (pingcap#57898) (pingcap#57935)
* planner: convert cartesian semi join with other nulleq condition to cr…
* planner: fix idxMergePartPlans forget to deal with RootTaskConds (pingcap#585…
* domain: change some stats log level as WARN (pingcap#58316) (pingcap#58555)
* planner: quickly get total count from index/column (pingcap#58365) (pingcap#58431)
* planner, expr: eval readonly user var during plan phase (pingcap#54462) (pingcap#58540)
* metrics: add col/idx name(s) for BackfillProgressGauge and BackfillTot…
* br: refactor test to use wait checkpoint method (pingcap#57612) (pingcap#58498)
* executor: reuse chunk in hash join v2 during restoring (pingcap#56936) (pingcap#58018)
* executor: fix goroutine leak when exceed quota in hash agg (pingcap#58078) (pingcap#58462)
* copr: fix the issue that busy threshold may redirect batch copr to fol…
* statistics: skip non-exicted table when to init stats (pingcap#58381) (pingcap#58394)
* planner: fix incorrectly using the schema for plan cache (pingcap#57964) (pingcap#58090)
* *: use DDL subscriber updating stats meta (pingcap#57872) (pingcap#58387)
* planner, runtime_filter: Remove redundant logs whose meaning can be di…
* statistics: remove dead code (pingcap#58412) (pingcap#58442)
* planner: Use/force to apply prefer range scan (pingcap#56928) (pingcap#58444)
* statistics: gc the statistics correctly after drop the database (pingcap#5730…
* ddl: Fixed partition interval from DayMinute to just Minute. (pingcap#57738) (pingcap#58019)
* executor: Enlarge the timeout for fetching TiFlash system tables (pingcap#579
@terry1purcell terry1purcell deleted the forced branch August 15, 2025 20:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
affects-8.5 This bug affects the 8.5.x(LTS) versions. approved lgtm needs-cherry-pick-release-8.5 Should cherry pick this PR to release-8.5 branch. ok-to-test Indicates a PR is ready to be tested. release-note-none Denotes a PR that doesn't merit a release note. sig/planner SIG: Planner size/XL Denotes a PR that changes 500-999 lines, ignoring generated files.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Planner avoid table scans for pseudo stats if index scan exists
6 participants