System information
| Type |
Version/Name |
| Distribution Name |
Debian and others |
| Distribution Version |
|
| Kernel Version |
|
| Architecture |
|
| OpenZFS Version |
|
Describe the problem you're observing
We've seen an uptick in zed_synchronous_zedlet ZTS failures. Example:
https://github.com/openzfs/zfs/actions/runs/21771045624/job/62999810149?pr=18161
Describe how to reproduce the problem
Run ZTS in CI
Include any warning/errors/backtraces from the system logs
Looks like it times out waiting for a trim_finish event after a pool trim.
[2026-02-07T02:32:42.074094] Test (Linux): /usr/share/zfs/zfs-tests/tests/functional/events/zed_synchronous_zedlet (run as root) [00:11] [FAIL]
02:32:30.25 ASSERTION: Verify ZED synchronous zedlets work as expected
02:32:30.25 SUCCESS: truncate -s 100M /var/tmp/vdev-file-sync-zedlet
02:32:30.31 SUCCESS: zpool create testpool2 /var/tmp/vdev-file-sync-zedlet
02:32:30.34 SUCCESS: zpool scrub -w testpool2
02:32:30.40 cleared 6522 events
02:32:30.40 SUCCESS: zpool events -c
02:32:30.41 SUCCESS: chmod +x /var/tmp/zed/scrub_start-async.sh
02:32:30.41 SUCCESS: chmod +x /var/tmp/zed/scrub_start-slow.sh
02:32:30.42 SUCCESS: chmod +x /var/tmp/zed/scrub_finish-sync-slow.sh
02:32:30.42 SUCCESS: chmod +x /var/tmp/zed/trim_start-async.sh
02:32:30.44 NOTE: Starting ZED
02:32:30.44 SUCCESS: truncate -s 0 /var/tmp/zed/zed.debug.log
02:32:30.44 SUCCESS: eval zed -vF -d /var/tmp/zed -P /var/tmp/constrained_path.IwSDc2 -s /var/tmp/zed/state -j 1 2>/var/tmp/zed/zed.log &
02:32:30.45 SUCCESS: zed_start
02:32:30.48 SUCCESS: zpool scrub -w testpool2
02:32:30.52 SUCCESS: zpool trim -w testpool2
02:32:40.57 file_wait_event exceeded 10.041 seconds
02:32:40.57 ERROR: file_wait_event /var/tmp/zed/zed.debug.log sysevent\.fs\.zfs\.trim_finish 10 exited 1
System information
Describe the problem you're observing
We've seen an uptick in
zed_synchronous_zedletZTS failures. Example:https://github.com/openzfs/zfs/actions/runs/21771045624/job/62999810149?pr=18161
Describe how to reproduce the problem
Run ZTS in CI
Include any warning/errors/backtraces from the system logs
Looks like it times out waiting for a
trim_finishevent after a pool trim.