Project

General

Profile

Actions
Copy link

Bug #49056

closed

faulty behaviour running ceph orch apply mds with missing fsname

Added by Kenneth Waegeman over 3 years ago. Updated about 3 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Category:
orchestrator
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
pacific
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
Ceph - v15.2.8
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

On 15.2.8 , I accidently ran `ceph orch apply mds label:mds`, so without the <fsname>.

This did not give an error message, but instead created these daemons:

# ceph orch ps
NAME                          HOST                STATUS         REFRESHED  AGE  VERSION    IMAGE NAME                   IMAGE ID      CONTAINER ID  
...
mds.label:mds.mds2801.gfahpt  mds2801.banette.os  error          4m ago     9m   <unknown>  docker.io/ceph/ceph:v15.2.8  <unknown>     <unknown>     
mds.label:mds.mds2801.wrmlmb  mds2801.banette.os  error          4m ago     9m   <unknown>  docker.io/ceph/ceph:v15.2.8  <unknown>     <unknown>     
mds.label:mds.mds2801.xpmusg  mds2801.banette.os  error          4m ago     9m   <unknown>  docker.io/ceph/ceph:v15.2.8  <unknown>     <unknown>   
...

I tried removing them (ceph orch daemon rm mds.label:mds.mds2803.aeutth) but that starts new ones, and this also doesnt work:

[root@mds2803 ~]# ceph orch apply mds mds.label:mds 0
Error EINVAL: num/count must be > 1

Any other way to get rid of these?

Thank you!

Kenneth

Actions
Copy link

Also available in: Atom PDF