Skip to content

release-23.2: colexecerror: avoid debug.Stack in CatchVectorizedRuntimeError #123502

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 4 commits into from
May 6, 2024
Merged

release-23.2: colexecerror: avoid debug.Stack in CatchVectorizedRuntimeError #123502

merged 4 commits into from
May 6, 2024

Conversation

michae2
Copy link
Collaborator

@michae2 michae2 commented May 2, 2024

Backport 4/4 commits from #123277 on behalf of @michae2.

/cc @cockroachdb/release

See individual commits for details.

Benchmarks before and after the change:

                                                                    │ /tmp/tmp.PrzUglo6GH/bench.catch~3 │    /tmp/tmp.PrzUglo6GH/bench.catch     │
                                                                    │              sec/op               │    sec/op      vs base                 │
CatchVectorizedRuntimeError/noError-12                                                    0.4033n ± ∞ ¹   0.4224n ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/expected-12                                                 40831.00n ± ∞ ¹    98.55n ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/storage-12                                                  41302.00n ± ∞ ¹    99.54n ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/contextCanceled-12                                          40947.00n ± ∞ ¹    99.78n ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/internalWithCode-12                                          41088.0n ± ∞ ¹    474.6n ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/internal-12                                                  41032.0n ± ∞ ¹    943.0n ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/runtime-12                                                    37.094µ ± ∞ ¹    1.091µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/noError-12                                         43.83µ ± ∞ ¹    43.63µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/expectedWithCode-12                              1234.24µ ± ∞ ¹    48.27µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/expectedAssertion-12                             1182.11µ ± ∞ ¹    88.12µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/internalAssertion-12                             1211.83µ ± ∞ ¹    77.58µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/internalIndexOutOfRange-12                       1227.76µ ± ∞ ¹    82.40µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/internalDivideByZero-12                          1224.45µ ± ∞ ¹    81.01µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/contextCanceled-12                               1265.30µ ± ∞ ¹    44.83µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/noError-12                                        43.78µ ± ∞ ¹    42.83µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/expectedWithCode-12                             1210.98µ ± ∞ ¹    48.41µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/expectedAssertion-12                            1187.21µ ± ∞ ¹    84.70µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/internalAssertion-12                            1224.62µ ± ∞ ¹    73.85µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/internalIndexOutOfRange-12                      1224.56µ ± ∞ ¹    81.39µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/internalDivideByZero-12                         1259.05µ ± ∞ ¹    80.63µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/contextCanceled-12                              1262.44µ ± ∞ ¹    47.13µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/noError-12                                        44.98µ ± ∞ ¹    43.53µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/expectedWithCode-12                             1231.37µ ± ∞ ¹    48.65µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/expectedAssertion-12                            1201.74µ ± ∞ ¹    97.23µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/internalAssertion-12                            1249.63µ ± ∞ ¹    80.46µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/internalIndexOutOfRange-12                      1257.86µ ± ∞ ¹    77.83µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/internalDivideByZero-12                         1251.79µ ± ∞ ¹    82.22µ ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=600/contextCanceled-12                              1271.82µ ± ∞ ¹    51.19µ ± ∞ ¹        ~ (p=1.000 n=1) ²
geomean                                                                                    243.2µ          13.09µ        -94.62%
¹ need >= 6 samples for confidence interval at level 0.95
² need >= 4 samples to detect a difference at alpha level 0.05

                                                                    │ /tmp/tmp.PrzUglo6GH/bench.catch~3 │    /tmp/tmp.PrzUglo6GH/bench.catch     │
                                                                    │               B/op                │     B/op       vs base                 │
CatchVectorizedRuntimeError/noError-12                                                      0.000 ± ∞ ¹     0.000 ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/expected-12                                                   9834.00 ± ∞ ¹     32.00 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/storage-12                                                    8880.00 ± ∞ ¹     32.00 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/contextCanceled-12                                            8840.00 ± ∞ ¹     40.00 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/internalWithCode-12                                           9.586Ki ± ∞ ¹   1.511Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/internal-12                                                  10.479Ki ± ∞ ¹   2.961Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/runtime-12                                                   10.964Ki ± ∞ ¹   2.656Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/noError-12                                        81.75Ki ± ∞ ¹   81.71Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/expectedWithCode-12                              143.04Ki ± ∞ ¹   85.28Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/expectedAssertion-12                              362.7Ki ± ∞ ¹   304.5Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/internalAssertion-12                              304.2Ki ± ∞ ¹   245.3Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/internalIndexOutOfRange-12                        313.0Ki ± ∞ ¹   267.2Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/internalDivideByZero-12                           308.3Ki ± ∞ ¹   260.7Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/contextCanceled-12                               128.96Ki ± ∞ ¹   80.60Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/noError-12                                       81.21Ki ± ∞ ¹   81.25Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/expectedWithCode-12                             136.02Ki ± ∞ ¹   84.99Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/expectedAssertion-12                             362.0Ki ± ∞ ¹   302.9Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/internalAssertion-12                             291.4Ki ± ∞ ¹   244.5Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/internalIndexOutOfRange-12                       317.1Ki ± ∞ ¹   267.0Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/internalDivideByZero-12                          308.4Ki ± ∞ ¹   260.4Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/contextCanceled-12                              145.34Ki ± ∞ ¹   81.73Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/noError-12                                       81.33Ki ± ∞ ¹   81.24Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/expectedWithCode-12                             137.49Ki ± ∞ ¹   85.05Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/expectedAssertion-12                             364.7Ki ± ∞ ¹   305.8Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/internalAssertion-12                             318.4Ki ± ∞ ¹   244.5Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/internalIndexOutOfRange-12                       315.8Ki ± ∞ ¹   265.4Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/internalDivideByZero-12                          310.9Ki ± ∞ ¹   260.2Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/contextCanceled-12                              133.20Ki ± ∞ ¹   80.53Ki ± ∞ ¹        ~ (p=1.000 n=1) ³
geomean                                                                                               ⁴                  -61.19%               ⁴
¹ need >= 6 samples for confidence interval at level 0.95
² all samples are equal
³ need >= 4 samples to detect a difference at alpha level 0.05
⁴ summaries must be >0 to compute geomean

                                                                    │ /tmp/tmp.PrzUglo6GH/bench.catch~3 │    /tmp/tmp.PrzUglo6GH/bench.catch    │
                                                                    │             allocs/op             │  allocs/op    vs base                 │
CatchVectorizedRuntimeError/noError-12                                                      0.000 ± ∞ ¹    0.000 ± ∞ ¹        ~ (p=1.000 n=1) ²
CatchVectorizedRuntimeError/expected-12                                                    36.000 ± ∞ ¹    3.000 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/storage-12                                                     11.000 ± ∞ ¹    3.000 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/contextCanceled-12                                             11.000 ± ∞ ¹    4.000 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/internalWithCode-12                                             35.00 ± ∞ ¹    37.00 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/internal-12                                                     46.00 ± ∞ ¹    59.00 ± ∞ ¹        ~ (p=1.000 n=1) ³
CatchVectorizedRuntimeError/runtime-12                                                      58.00 ± ∞ ¹    51.00 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/noError-12                                          838.0 ± ∞ ¹    838.0 ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=12/expectedWithCode-12                                1110.0 ± ∞ ¹    950.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/expectedAssertion-12                               1.736k ± ∞ ¹   1.581k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/internalAssertion-12                               1.285k ± ∞ ¹   1.228k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/internalIndexOutOfRange-12                         1.418k ± ∞ ¹   1.425k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/internalDivideByZero-12                            1.413k ± ∞ ¹   1.411k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=12/contextCanceled-12                                  880.0 ± ∞ ¹    865.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/noError-12                                         836.0 ± ∞ ¹    836.0 ± ∞ ¹        ~ (p=1.000 n=1) ²
SQLCatchVectorizedRuntimeError/conns=240/expectedWithCode-12                               1012.0 ± ∞ ¹    949.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/expectedAssertion-12                              1.744k ± ∞ ¹   1.572k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/internalAssertion-12                              1.201k ± ∞ ¹   1.228k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/internalIndexOutOfRange-12                        1.518k ± ∞ ¹   1.424k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/internalDivideByZero-12                           1.429k ± ∞ ¹   1.411k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=240/contextCanceled-12                                 993.0 ± ∞ ¹    865.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/noError-12                                         838.0 ± ∞ ¹    837.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/expectedWithCode-12                               1038.0 ± ∞ ¹    951.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/expectedAssertion-12                              1.782k ± ∞ ¹   1.594k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/internalAssertion-12                              1.471k ± ∞ ¹   1.231k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/internalIndexOutOfRange-12                        1.476k ± ∞ ¹   1.417k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/internalDivideByZero-12                           1.471k ± ∞ ¹   1.411k ± ∞ ¹        ~ (p=1.000 n=1) ³
SQLCatchVectorizedRuntimeError/conns=600/contextCanceled-12                                 960.0 ± ∞ ¹    867.0 ± ∞ ¹        ~ (p=1.000 n=1) ³
geomean                                                                                               ⁴                 -18.86%               ⁴
¹ need >= 6 samples for confidence interval at level 0.95
² all samples are equal
³ need >= 4 samples to detect a difference at alpha level 0.05
⁴ summaries must be >0 to compute geomean

Fixes: #123235

Release note (performance improvement): Make error handling in the vectorized execution engine much cheaper. This should help avoid bad metastable regimes perpetuated by statement timeout handling consuming all CPU time, leading to more statement timeouts.

Co-authored-by: Drew Kimball [email protected]

Release justification: fix for serious performance issue.

@michae2 michae2 requested review from mgartner and rafiss May 2, 2024 19:07
@michae2 michae2 requested review from a team as code owners May 2, 2024 19:07
@blathers-crl blathers[crl]
Copy link

blathers-crl bot commented May 2, 2024

Thanks for opening a backport.

Please check the backport criteria before merging:

  • Backports should only be created for serious
    issues     or test-only changes.
  • Backports should not break backwards-compatibility.
  • Backports should change as little code as possible.
  • Backports should not change on-disk formats or node communication protocols.
  • Backports should not add new functionality (except as defined
    here).
  • Backports must not add, edit, or otherwise modify cluster versions; or add version gates.
  • All backports must be reviewed by the owning areas TL and one additional
    TL. For more information as to how that review should be conducted, please consult the backport
    policy    .
If your backport adds new functionality, please ensure that the following additional criteria are satisfied:
  • There is a high priority need for the functionality that cannot wait until the next release and is difficult to address in another way.
  • The new functionality is additive-only and only runs for clusters which have specifically “opted in” to it (e.g. by a cluster setting).
  • New code is protected by a conditional check that is trivial to verify and ensures that it only runs for opt-in clusters. State changes must be further protected such that nodes running old binaries will not be negatively impacted by the new state (with a mixed version test added).
  • The PM and TL on the team that owns the changed code have signed off that the change obeys the above rules.
  • Your backport must be accompanied by a post to the appropriate Slack
    channel (#db-backports-point-releases or #db-backports-XX-X-release) for awareness and discussion.

Also, please add a brief release justification to the body of your PR to justify this
backport.

@blathers-crl blathers-crl bot added the backport Label PR's that are backports to older release branches label May 2, 2024
@cockroach-teamcity
Copy link
Member

This change is Reviewable

@michae2 michae2 removed request for a team May 2, 2024 19:19
rafiss
rafiss approved these changes May 2, 2024
View reviewed changes
Copy link
Collaborator

@rafiss rafiss left a comment

Choose a reason for hiding this comment

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

Reviewed 7 of 7 files at r1.
Reviewable status: :shipit: complete! 0 of 0 LGTMs obtained (waiting on @mgartner)

DrewKimball
DrewKimball approved these changes May 2, 2024
View reviewed changes
Copy link
Collaborator

@DrewKimball DrewKimball left a comment

Choose a reason for hiding this comment

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

:lgtm:

Reviewed 7 of 7 files at r1, 1 of 1 files at r2, 1 of 1 files at r3, 1 of 1 files at r4, all commit messages.
Reviewable status: :shipit: complete! 1 of 0 LGTMs obtained (waiting on @mgartner)

mgartner
mgartner approved these changes May 3, 2024
View reviewed changes
Copy link
Collaborator

@mgartner mgartner left a comment

Choose a reason for hiding this comment

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

:lgtm:

Reviewed 7 of 7 files at r1, 1 of 1 files at r2, 1 of 1 files at r3, 1 of 1 files at r4, all commit messages.
Reviewable status: :shipit: complete! 2 of 0 LGTMs obtained (waiting on @michae2)

@michae2
Copy link
Collaborator Author

michae2 commented May 3, 2024 •
edited
Loading

Test failure looks like an unrelated flake, so I'll go ahead.

Edit: hmm, actually going to do a no-op amend to try and kick Essential CI again.

michae2 added 4 commits May 3, 2024 12:28
@michae2
colexecerror: add benchmarks for CatchVectorizedRuntimeError
52f3d60 
Informs: cockroachdb#123235

Release note: None
@michae2
colexecerror: check for expected errors before examining stack
0a21918 
Informs: cockroachdb#123235

Release note: None
@michae2
colexecerror: check for internal errors before examining stack
9a1f70b 
Informs: cockroachdb#123235

Release note: None
@michae2
colexecerror: use runtime.Callers instead of debug.Stack
dff8312 
Fixes: cockroachdb#123235

Release note (performance improvement): Make error handling in the
vectorized execution engine much cheaper. This should help avoid bad
metastable regimes perpetuated by statement timeout handling consuming
all CPU time, leading to more statement timeouts.
@michae2 michae2 force-pushed the backport23.2-123277 branch from 2f61d52 to dff8312 Compare May 3, 2024 19:28
@michae2 michae2 merged commit a4263db into cockroachdb:release-23.2 May 6, 2024
5 of 6 checks passed
@michae2 michae2 deleted the backport23.2-123277 branch May 6, 2024 17:22
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@mgartner mgartner mgartner approved these changes

@rafiss rafiss rafiss approved these changes

@DrewKimball DrewKimball DrewKimball approved these changes

Assignees
No one assigned
Labels
backport Label PR's that are backports to older release branches
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
5 participants
@michae2 @cockroach-teamcity @mgartner @rafiss @DrewKimball