the "use crs to select least loaded node when starting an HA service" is a great feature, but I certainly dont want that to apply to lxc containers since they have to be rebooted on migration; is there an option to limit this to virtual machines?
Hi all,
I'm testing new setup where I have 3x storage boxes(with multiples drives) and 6x compute boxes (with two drives + more RAM) all are part of same cluster.
I installed ceph on 3x storage nodes and added all the free drives from it as OSDs. ceph is up and running and I can use mounted...