Hi all,
I have a 3 nodes CEPH cluster, each node with 4 OSD and 1 monitor and replication across hosts defined in the crushmap. 2 pools (1 size=3, min-size=2 , 1 size=2 min-size 2)
One of the nodes failed (hardware failure) so I removed it from the crush map and change replication from host to osd.
After reconstruction cluster health is ok now (with a warning for the monitor down)
While I managed to restore the 3rd node is there a way to cover a 2nd node failure, leaving just 1 node in the cluster (disk space used does not exceed 1 node capacity).
I assume i will have to set up an external monitor, but what else ?
Olivier
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
