Skip to content
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.

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

Disable failing ci builders #7206

Merged
merged 10 commits into from
Dec 14, 2023
Merged

Disable failing ci builders #7206

merged 10 commits into from
Dec 14, 2023

Conversation

kiburtse
Copy link
Contributor

@kiburtse kiburtse commented Dec 13, 2023

What, How & Why?

  • Disable builders testing faulty network - it was not supposed to run since it's not yet ready, and 'disable' tag doesn't work as expected
  • Also, don't run core/sync internal test on new macos arm64 with tsan since it's not stable with apple clang 13 TSAN issue: ThreadSanitizer:DEADLYSIGNAL #7185
  • Introduce new builders for asan/tsan on ubuntu arm64

☑️ ToDos

  • 📝 Changelog update
  • 🚦 Tests (or not relevant)
  • C-API, if public C++ API changed
  • bindgen/spec.yml, if public C++ API changed

Copy link

coveralls-official bot commented Dec 13, 2023

Pull Request Test Coverage Report for Build kirill.burtsev_116

  • 0 of 0 changed or added relevant lines in 0 files are covered.
  • 21 unchanged lines in 7 files lost coverage.
  • Overall coverage increased (+0.03%) to 91.709%

Files with Coverage Reduction New Missed Lines %
src/realm/sync/noinst/server/server_history.cpp 2 67.94%
src/realm/uuid.cpp 2 97.06%
src/realm/sync/noinst/client_impl_base.cpp 3 85.28%
src/realm/sync/noinst/server/server.cpp 3 76.11%
src/realm/util/file.cpp 3 81.52%
test/fuzz_group.cpp 3 48.39%
src/realm/alloc_slab.cpp 5 92.93%
Totals Coverage Status
Change from base Build 1915: 0.03%
Covered Lines: 232238
Relevant Lines: 253233

💛 - Coveralls

@kiburtse
Copy link
Contributor Author

Apparently, ubuntu arm64 is even less ready for sanitizers...

What is the list of distros we have for evergreen runners? Is there ubuntu2204-arm64-large for example?

@kiburtse
Copy link
Contributor Author

ok.. using ubuntu2204-arm64 for tsan opens whole can of worms including #7083 #6193 #6194. Let's skip it for now.

macos1300-arm64 for tsan can't find the compiler. Any ideas why?

Comment on lines +1571 to +1584
- name: ubuntu2204-arm64-asan
display_name: "Ubuntu 22.04 ARM64 (ASAN)"
run_on: ubuntu2204-arm64-large
expansions:
cmake_url: "https://s3.amazonaws.com/static.realm.io/evergreen-assets/cmake-3.20.3-linux-aarch64.tar.gz"
cmake_bindir: "./cmake_binaries/bin"
python3: "/opt/mongodbtoolchain/v3/bin/python3"
use_system_openssl: On
fetch_missing_dependencies: On
cmake_build_type: Debug
enable_asan: On
tasks:
- name: compile_test

Copy link
Contributor

Choose a reason for hiding this comment

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

So you got ubuntu2204-arm64-asan to pass by suppressing the messages around adjtime - nice!

Copy link
Contributor Author

Choose a reason for hiding this comment

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

not quite, this is asan, not tsan. The problem with adjtime was a new one after #6911. It also fails builds from time to time. TSAN on ubuntu on arm64 still needs a lot of fixes. The test build on ubuntu2204-arm64-large showed a few hundred issues, i didn't yet look into them apart from linked open ones.

Copy link
Contributor

@michael-wb michael-wb left a comment

Choose a reason for hiding this comment

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

Couple of questions.

Don't forget to add the no-changelog label if you aren't planning to add a changelog entry.

Comment on lines +1730 to +1731
# FIXME: tsan is not stable on arm64, fails often with internal errors
# - name: compile_test
Copy link
Contributor

Choose a reason for hiding this comment

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

Was this for macOS as well? I thought it was only for ubuntu2204...

Copy link
Contributor Author

@kiburtse kiburtse Dec 14, 2023

Choose a reason for hiding this comment

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

this is for #7185 - this failure is presumably core_tests and sync_tests specific only on macos 11 arm64. I haven't seen this failures myself on macos 13 arm64, so i hope if we move this builder on macos1300-arm64 then we can reenable whole test suite.

ubuntu on arm64 reports different issues but consistently and for every test suite we have from what i can tell

- name: compile_test
# FIXME: tsan is not stable on arm64, fails often with internal errors
# - name: compile_test
- name: compile_test_object_store
Copy link
Contributor

Choose a reason for hiding this comment

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

Do you know if the sync-tests pass on this platform?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

they do, but way too often they fail #7185

Copy link
Member

@nicola-cab nicola-cab left a comment

Choose a reason for hiding this comment

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

@kiburtse you are disabling the extra builders for TSAN only for evergreen as far as I can see. In relation to my comment for this issue, #7185 ... we could potentially get away, since we are releasing our binaries via jenkins (for now). If we are not touching Jenkins, we should be fine.
Other than this, it looks good to me.

@kiburtse
Copy link
Contributor Author

@kiburtse you are disabling the extra builders for TSAN only for evergreen as far as I can see. In relation to my comment for this issue, #7185 ... we could potentially get away, since we are releasing our binaries via jenkins (for now). If we are not touching Jenkins, we should be fine. Other than this, it looks good to me.

well, tsan on macos arm64 was a new builder, apparently it's not very stable. Even compiler on this macos1100-arm64 is not the latest one from AppleClang 13, which should be 13.0.1. I see no point in fighting it and tolerating this failures. It was well over 10s of failures in the last month along. Our ci is constantly red, it's hard to keep track of all these failures. So i'd rather disable partially for now, and resume with macos1300 - it should be better i hope.

@nicola-cab nicola-cab self-requested a review December 14, 2023 17:37
@kiburtse kiburtse merged commit e621241 into master Dec 14, 2023
33 of 35 checks passed
@kiburtse kiburtse deleted the kb/disable_failing_ci_builders branch December 14, 2023 17:44
@github-actions github-actions bot locked as resolved and limited conversation to collaborators Mar 21, 2024
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants