Hi,
could you be hitting the bug from [1]? Watch out for segfaults in dmesg.
Since a couple of days we see random OSDs with a segfault from
safe_timer. We didn't update any packages for months.
Regards
[1] https://tracker.ceph.com/issues/23352
Zitat von Rudenko Aleksandr <ARudenko@xxxxxxx>:
Hi, guys.
After upgrade on Luminous i see:
Monitor daemon marked osd.xx down, but it is still running
this happens 3-5 times a day on different OSDs.
I spent a lot of time on debug but i haven’t found problem.
Network works perfectly. CPU, network and disk utilization is low.
Memory is enough.
Maybe deep-scrub, but we have following config:
osd scrub sleep = 0.2
osd scrub chunk min = 1
osd scrub chunk max = 2
and we didn’t see OSD flapping on Hammer and Jewel during scrub.
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
