Yu Watanabe
Yu Watanabe
Interestingly, this is not a bug in udev or TEST-64. This is caused by that the VM is restarted after poweroff and the same test is started twice, and TEST-64-UDEV-STORAGE-mdadm_lvm...
``` [ 22.667697] systemd[1]: TEST-64-UDEV-STORAGE-mdadm_lvm.service: Consumed 7.246s CPU time. [ 22.667706] systemd[1]: Doing "poweroff" action instead of an "exit" emergency action. [ 22.667710] systemd[1]: Powering off: unit TEST-64-UDEV-STORAGE-mdadm_lvm.service succeeded [...
@DaanDeMeyer Do you have any idea?
Not checked, but potential another failure in TEST-46-HOMED: https://github.com/systemd/systemd/actions/runs/16554966342/job/46815132261?pr=38372
Yeah, the VM is rebooted: ``` $ journalctl --file TEST-46-HOMED-1.journal --list-boots IDX BOOT ID FIRST ENTRY LAST ENTRY -1 1edd3b321fce4547b76956db867ec664 Mon 2025-07-28 05:42:10 JST Mon 2025-07-28 05:43:35 JST 0 5bb99456b57d4a9e943b4977b0d7e996...
So, the issue should be in one of the following: - qemu - vmspawn - mkosi - systemd-shutdown
Another failure: https://github.com/systemd/systemd/actions/runs/16575712734/job/46879378581?pr=38373
Currently the issue is found only on Arch and Fedora rawhide. So, I guess this is a bug in qemu. Though, still we have only 4 samples.
The workaround seems to not enough: https://github.com/systemd/systemd/actions/runs/16729546470/job/47354007658 Also, the issue is triggered even on CentOS 9.
workaround fix: https://github.com/systemd/systemd/pull/38478