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

Revert "refactor(VirtualFile::crashsafe_overwrite): avoid Handle::block_on in callers" #6765

Conversation

problame
Copy link
Contributor

Reverts #6731

On high tenant count Pageservers in staging, memory and CPU usage shoots to 100% with this change. (NB: staging currently has tokio-epoll-uring enabled)

Will analyze tomorrow.

https://neondb.slack.com/archives/C03H1K0PGKH/p1707933875639379?thread_ts=1707929541.125329&cid=C03H1K0PGKH

@problame problame requested a review from a team as a code owner February 14, 2024 18:22
@problame problame requested review from arpad-m and jcsp and removed request for a team and arpad-m February 14, 2024 18:22
@jcsp jcsp enabled auto-merge (squash) February 14, 2024 18:24
Copy link

2436 tests run: 2318 passed, 0 failed, 118 skipped (full report)


Flaky tests (2)

Postgres 16

Postgres 14

  • test_timeline_deletion_with_files_stuck_in_upload_queue: debug

Code coverage (full report)

  • functions: 56.0% (12864 of 22978 functions)
  • lines: 82.6% (69488 of 84173 lines)

The comment gets automatically updated with the latest test results
45f2d6a at 2024-02-14T19:05:37.879Z :recycle:

@jcsp jcsp merged commit 024372a into main Feb 14, 2024
49 checks passed
@jcsp jcsp deleted the revert-6731-problame/integrate-tokio-epoll-uring/write-path/crashsafe_overwrite/switch-to-spawn-blocking branch February 14, 2024 19:17
problame added a commit that referenced this pull request Feb 23, 2024
…dle::block_on in callers"" (#6775)

Reverts #6765 , bringing back #6731

We concluded that #6731 never was the root cause for the instability in
staging.
More details:
https://neondb.slack.com/archives/C033RQ5SPDH/p1708011674755319

However, the massive amount of concurrent `spawn_blocking` calls from
the `save_metadata` calls during startups might cause a performance
regression.
So, we'll merge this PR here after we've stopped writing the metadata
#6769).
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants