You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We had some scheduler/threading issues from 27-09-2024 21:00 (CET) untill 28-09-2024 09:00 (CET) leading to high error-rates (~25%) on uncached images.
The automatic reloading of nginx config every few hours seems to have worsened the problem, leading to stalled or unresponsive threads.
We didn't catch this issue in earlier testing of some libvips code changes, as the reloading nginx under load is not part of automated tests.
We're sorry for the troubles caused, the partial outage surprised us as well.
What I've noticed since this morning is that for certain images (most likely the new, not cached ones), the connections always time out.
The text was updated successfully, but these errors were encountered: