On Thu, May 12, 2016 at 2:54 PM, Stephen Mercier <stephen.mercier@xxxxxxxxxxxx> wrote:
Thank you very much for the thorough explanation. What you described was one of the ways I was interpreting this.Now, out of curiosity, if I did:rule replicated_rack {ruleset 0type replicatedmin_size 1max_size 10step take defaultstep choose firstn 2 type rackstep chooseleaf firstn 2 type hoststep emit}Would that choose the 2 racks, and then 2 osds from each host, making sure that in a 3x replication scenario, I would ensure that all 3 reps would be on different servers as well? Just want to make sure I'm following the logic correctly before I pull any triggers.
Yep!
On Thu, May 12, 2016 at 2:57 PM, Stephen Mercier <stephen.mercier@xxxxxxxxxxxx> wrote:
Also, when you say new enough for the overselection hack, how new are we talking?!?!
Not very. I'm honestly not sure and you're best off experimenting — all the current userspace stuff is fine; there might still be a kernel client in LTS releases which won't be happy with it.
-Greg
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com