I was having horrible problems getting my test ceph clusterj reinitialized.
All kinds of annoying things were happening.
including things like getting differing output from
ceph orch device ls
vs
ceph device ls
Being new-ish to ceph, i was going nuts, wondering what kind of init options I was
missing.
Turns out, nothing i was doing was wrong, per se.
I had ended up with differing container versions.
Even after doing "cephadm rm-cluster", the old versions were sticking.
The really annoying thing is, the difference was tiny.
15.2.8 on the master.
15.2.5 on other nodes.
but device related things were failing, with
2020-12-21 09:37:09,134 INFO /bin/podman:stderr ceph-volume inventory: error: unrecognized
arguments: --filter-for-batch
in /var/log/ceph/cephadm.log
Sighhh...
To save anyone else some research, the end-user fix is:
run
ceph orch upgrade start --ceph-version 15.2.8
--
Philip Brown| Sr. Linux System Administrator | Medata, Inc.
5 Peters Canyon Rd Suite 250
Irvine CA 92606
Office 714.918.1310| Fax 714.918.1325
pbrown(a)medata.com|
www.medata.com