Hello ! Since we have upgraded our Ceph cluster from Ceph Pacific 16.2.15 to Ceph 18.2.2. We can see more requests in timeout. Especially PUT from backup tools, Nexus Repository... Een with 10min timeouts on haproxy for the server backends. We are upgrading Ubuntu from 18.04 to 22.04 to be able to upgrade to Ceph 18.2.4, but not finished yet. With debug log level, I can see only one thing concerning a never ending request. Here are the last logs : Dec 09 12:50:20 server-xxx bash[876499]: debug 2024-12-09T11:50:20.189+0000 7f5af9890700 10 req 7017125235642647451 0.004000126s s3:put_obj setting object write_tag=3e3b0a2b-1864-4923-93a6-237a13b51594.81430478.7017125235642647451 Dec 09 12:50:20 server-xxx bash[876499]: debug 2024-12-09T11:50:20.189+0000 7f5af9890700 20 req 7017125235642647451 0.004000126s s3:put_obj bucket index object: MYREGION.rgw.buckets.index:.dir.3e3b0a2b-1864-4923-93a6-237a13b51594.80569777.1.5 Dec 09 12:50:20 server-xxx bash[876499]: debug 2024-12-09T11:50:20.341+0000 7f5a83fa5700 20 req 7017125235642647451 0.156004906s s3:put_obj WARNING: blocking librados call I wonder if that last line is the culprit, end if the prcess ends or stay blocked indefinitelly... --Gilles _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |