ci/nixpkgs-vet: fix random errors#413239
Conversation
Every now and then, the nixpkgs-vet CI job currently fails with one of: error: creating symlink '/build/.local/share/nix/root/nix/var/nix/gcroots/profiles' -> '/build/.local/share/nix/root/nix/var/nix/profiles': File exists error: SQLite database '/build/.local/share/nix/root/nix/var/nix/db/db.sqlite' is busy It's hard to reproduce for me, so just taking a guess with the required changes.
github-actions
bot
added
6.topic: continuous integration
nix-owners
bot
requested review from
LeSuisse,
Mic92,
azuwis,
infinisil,
mweinelt,
philiptaron,
risicle and
zowoq
github-actions
bot
added
10.rebuild-darwin: 0
philiptaron
left a comment
philiptaron
left a comment
There was a problem hiding this comment.
I'm willing to be speculative here.
|
Successfully created backport PR for |
github-actions
bot
added
the
8.has: port to stable
|
Successfully created backport PR for |
|
Backport failed for Please cherry-pick the changes locally and resolve any conflicts. git fetch origin release-25.05
git worktree add -d .worktree/backport-413239-to-release-25.05 origin/release-25.05
cd .worktree/backport-413239-to-release-25.05
git switch --create backport-413239-to-release-25.05
git cherry-pick -x 3dd14d8a5176e7a790e6eeeab7bfebd49c4b8eab |
|
This seems to have happened after the fix, too: https://github.com/NixOS/nixpkgs/actions/runs/15426171510/job/43413501065 This was after merging this PR and shows the temp path used here. It seems like it happened less frequently, though, I had to go back quite a few failed workflows to find this. |
|
Unfortunate. I wonder if the shenanigans we use to invoke Nix as a tool vs. as a builder are the ones that trigger this unfortunate behavior. Because it's so inconsistent, it has to be a race somewhere with something, I would think. |
|
Here's the upstream Nix issue that causes this: NixOS/nix#2706 |
|
Proposed a fix in NixOS/nix#13368. |
|
... and merged already! Thank you @edolstra! |
2 participants
Every now and then, the nixpkgs-vet CI job currently fails with one of:
or
It's hard to reproduce for me, so just taking a guess with the required changes.
I bet somebody else seen those already and can tell whether my change here is likely to fix them.
Things done
Add a 👍 reaction to pull requests you find important.