On 1/26/19 10:24 PM, Kevin Olbrich wrote:
I just had the time to check again: even after removing the broken OSD, mgr still crashes. All OSDs are on and in. If I run "ceph balancer on" on a HEALTH_OK cluster, an optimization plan is generated and started. After some minutes all MGRs die. This is a major problem for me, as I still got that SSD OSD that is inbalanced and limiting the whole pools space.
Try to run mgr with `debug mgr = 4/5` and look to mgr log file. k _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
