Thank you Eugen for looking into it!
In short, it works. I'm using 16.2.10.
What I did wrong was to remove the OSD, which makes no sense.
Tony
From: Eugen Block
Sent: April 28, 2023 06:46 AM
To: ceph-users@ceph.io
Subject: [ceph-users] Re: import OSD
I chatted with Mykola who helped me get the OSDs back up. My test
cluster was on 16.2.5 (and still mostly is), after upgrading only the
MGRs to a more recent version (16.2.10) the activate command worked
successfully and the existing OSDs got back up. Not sure if that's a
bug or something
I found a small two-node cluster to test this on pacific, I can
reproduce it. After reinstalling the host (VM) most of the other
services are redeployed (mon, mgr, mds, crash), but not the OSDs. I
will take a closer look.
Zitat von Tony Liu :
Tried [1] already, but got error.
Created no
Hi,
Not sure what's missing.
Should OSD be removed, or removed with --replace, or untouched
before host reinstallation?
If you want to reuse existing OSDs why would you remove them? That's
the whole point of reusing them after installation.
Tried [1] already, but got error.
Created no
Tried [1] already, but got error.
Created no osd(s) on host ceph-4; already created?
The error is from [2] in deploy_osd_daemons_for_existing_osds().
Not sure what's missing.
Should OSD be removed, or removed with --replace, or untouched before host
reinstallation?
[1]