Some context. I have a small cluster running ubuntu 14.04 and giant ( now
hsmmer). I ran some updates everything was fine. Rebooted a node and a
drive must have failed as it no longer shows up.
I use --dmcrypt with ceph deploy and 5 osds per ssd journal. To do this I
created the ssd partitions already and pointed ceph-deploy towards the
partition for the journal.
This worked in giant without issue (I was able to zap the osd and redeploy
using the same journal all of the time). Now it seems to fail in hammer
stating that the partition exists and im using - - decrypt.
This raises a few questions.
1.) ceph osd start scripts must have a list of dm-crypt keys and uuids
somewhere as the init mounts the drives. Is this accessible? Normally
outside of ceph I've used crypt tab, how is ceph doing it?
2.) my ceph-deploy line is:
ceph-deploy osd --dmcrypt create ${host}:/dev/drive:/dev/journal_partition
I see that a variable in ceph-disk exists in and is set to false. Is this
what I would need to change to get this working again? Or is this set to
false for a reason?
3.) I see multiple references to journal_uuid in Sebastian Hans blog as
well as the mailing list when replacing a disk. I don't have this file,
and I'm assuming it's due to the - - dmcrypt flag. I also see 60
dmcrypt-keys in /etc/ceph/dmxrypt-keys but only 30 mapped devices. Are the
journals not using these keys at all?
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
