Not sure I mentioned, but I am on CentOS 7.3.
I tried to insert a sleep in ExecStartPre in /usr/lib/systemd/system/ceph-osd@.service but apparently all ceph-osd are started (and retried) at the same time.
I finally noticed that the simple
ceph-disk activate <ceph_data_partition>
is sufficient to recover the OSD.
Questions:
= why am I not able to restart the OSD via
systemctl restart ceph-osd@##.service
whereas ceph-disk activates magically works?
= (off-topic) I also see systemd complaining about OSD## which
at some point existed on the host but later were reassigned to
another one. Tried to "systemctl stop/disable ceph-osd@##" but
those seem to reappear at boot... any idea how to fix this?
I could easily take care of the "OSD not activating at boot" with
something simple in rc.local, but I wonder whether someone is aware of a
cleaner solution.
Thanks!
Fulvio
smime.p7s
Description: S/MIME Cryptographic Signature
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
