Possible regression? Kernel cephfs >= 6.10 cpu hangup
Hello fellow ceph users, i've noticed that since i've upgraded to kernel version 6.10 i've been occasionally having system lockups with dmesg showing cephfs in the kernel trace. Due to the circumstances of the lockups, i've so far been unable to store these trace logs, but i will attempt to do so in a more isolated and controlled environment. The symptom seems to occur during file syncs (flushing), more specifically it has happened 2 times while saving a file from vim (create new file, flush to file, overwrite old file). Is anyone aware of changes that may have caused such a problem before i start digging deeper? Have a nice day -- Alex D. RedXen System & Infrastructure Administration https://redxen.eu/
Hello again, i've been able to extract the traces regarding this error and i'll provide more information: Mount options (excluding helper provided): 'recover_session=clean,fsc=XXXX' Kernel version: 6.10.2 Ceph versions: { "mon": { "ceph version 18.2.1 (e3fce6809130d78ac0058fc87e537ecd926cd213) reef (stable)": 2, "ceph version 18.2.2 (531c0d11a1c5d39fbfe6aa8a521f023abf3bf3e2) reef (stable)": 1 }, "mgr": { "ceph version 18.2.1 (e3fce6809130d78ac0058fc87e537ecd926cd213) reef (stable)": 2, "ceph version 18.2.2 (531c0d11a1c5d39fbfe6aa8a521f023abf3bf3e2) reef (stable)": 1 }, "osd": { "ceph version 18.2.1 (e3fce6809130d78ac0058fc87e537ecd926cd213) reef (stable)": 12, "ceph version 18.2.2 (531c0d11a1c5d39fbfe6aa8a521f023abf3bf3e2) reef (stable)": 6 }, "mds": { "ceph version 18.2.1 (e3fce6809130d78ac0058fc87e537ecd926cd213) reef (stable)": 4, "ceph version 18.2.2 (531c0d11a1c5d39fbfe6aa8a521f023abf3bf3e2) reef (stable)": 2 }, "overall": { "ceph version 18.2.1 (e3fce6809130d78ac0058fc87e537ecd926cd213) reef (stable)": 20, "ceph version 18.2.2 (531c0d11a1c5d39fbfe6aa8a521f023abf3bf3e2) reef (stable)": 10 } } Kernel log snippet: -- Alex D. RedXen System & Infrastructure Administration https://redxen.eu/
participants (1)
-
caskd