Thank you Stefan!
My problem was that the ruleset I built had the failure domain set to
rack, when I do not have any racks defined. I changed the failure domain
to host as this is just a home lab environment. I reverted the ruleset
on the pool, and it immediately started to recover and storage was
available to my virtual machines again. I then fixed the failure domain
and the ruleset works.
On 9/20/21 01:37, Stefan Kooman wrote:
On 9/20/21 07:51, Mr. Gecko wrote:
Hello,
I'll start by explaining what I have done. I was adding some new
storage in attempt to setup a cache pool according to
https://docs.ceph.com/en/latest/dev/cache-pool/ by doing the following.
1. I upgraded all servers in cluster to ceph 15.2.14 which put the
system into recovery for out of sync data.
2. I added 2 SSDs as OSDs to the cluster which immediately cause ceph
to balance onto the SSDs.
3. I added 2 new crush rules which map to SSD storage vs HDD storage.\
I guess this is were things go wrong. Have you tested the CRUSH rules
beforehand? To see if the right OSDs get mapped, or any at all.
I would revert the crush rule change for now to try get your PGs
active+clean.
If that works, than try to find out (with crushtool for example) why
the new CRUSH rule sets do not map the OSDs.
Gr. Stefan
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
