ceph: 18.2.0 -> 18.2.1 #281474
ceph: 18.2.0 -> 18.2.1 #281474
Conversation
ofborg
bot
added
10.rebuild-darwin: 1-10
10.rebuild-darwin: 1
10.rebuild-linux: 1-10
labels
|
Result of 13 packages built:
|
|
@ofborg build ceph ceph.passthru.tests |
There was a problem hiding this comment.
Changelog looks good, ofborg is happy (ran NixOS tests).
What's not working is aarch64-linux:
- https://github.com/NixOS/nixpkgs/pull/281474/checks?check_run_id=20962544866
- It worked for the 18.2.0 PR: https://github.com/NixOS/nixpkgs/pull/281858/checks?check_run_id=20823371445
The Python library watchfiles-0.21.0 seems to timeout in a test on this architecture:
error: builder for '/nix/store/kgy19r4gl6bcg1jjla9rvwlpgcfw77qj-python3.10-watchfiles-0.21.0.drv' failed with exit code 1;
last 10 log lines:
> path = os.getenv('WATCHFILES_TEST_PERMISSION_DENIED_PATH') or '/'
>
> > RustNotify([path], False, False, 0, True, True)
> E Failed: Timeout >10.0s
>
> tests/test_rust_notify.py:317: Failed
> =========================== short test summary info ============================
> FAILED tests/test_rust_notify.py::test_ignore_permission_denied - Failed: Timeout >10.0s
> ====== 1 failed, 150 passed, 2 skipped, 1 deselected in 80.39s (0:01:20) =======
> /nix/store/in5b9lna5p97wak30r8397645akl5092-stdenv-linux/setup: line 1582: pop_var_context: head of shell_variables not a function context
It does seem to build on Hydra with Python 3.11: https://hydra.nixos.org/build/247427896
I don't have an aarch64 builder so I can't easily investigate this.
I'm approving this PR under the assumption that this failure has nothing to do with Ceph.
|
@GrahamcOfBorg build ceph ceph.passthru.tests Just in case that test is flaky and works upon rebuild. |
|
Strangely, ceph builds for me on |
@fpletz Can you post the output? |
|
I force-pushed the PR to rebase on the merged PR #281858 with my Ceph GCC fixes. |
|
@GrahamcOfBorg build ceph ceph.passthru.tests |
|
Looks good to me. I can try to get at an aarch64-linux if we think there's truly a problem? |
|
By the way, I also see (racy) failures of ceph-mgr modules (also in prod, sadly) which seem related to the internal subinterpreter struggle, but worsened by pyo3 updates, but I've yet to see a complete test failure because of it. That may just be nondeterminism in the mgr module ordering though, I suppose. |
|
Ofborg now succeeds on https://github.com/NixOS/nixpkgs/pull/281474/checks?check_run_id=21247371804 I upgraded a test cluster from 16.2.7 to 18.2.1 based on this commit, so this is good enough to merge now.
@srhb Thanks for the info! I hope Ceph fixes this issue. Personally I enjoy that I saw the Ceph dashboard for the first time yesterday on my cluster on 18.2.1 :) |
|
We should backport this, right? |
I think so; I added the tag. |
|
Successfully created backport PR for |
Description of changes
Changelog: https://docs.ceph.com/en/latest/releases/reef/#v18-2-1-reef
Things done
nix.conf? (See Nix manual)sandbox = relaxedsandbox = truenix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage./result/bin/)Add a 馃憤 reaction to pull requests you find important.