On Wed, Dec 13, 2017 at 10:00 PM, Florent B <florent@xxxxxxxxxxx> wrote: > Hi, > > Trying to solve my problem of corrupted files on CephFS, I create a new > thread to talk about the warning "1 MDSs report slow requests" which > often occur when a process locks a file for a long time while others > processes asks for the lock too. > > Is it a wrong situation ? Does it really need to set cluster state to > HEALTH_WARN ? we don't need to set cluster unhealth state. created http://tracker.ceph.com/issues/22428 > For my example, here are my ops_in_flight & MDS cache dump related to a > file in this situation (see attachments). > > Thank you > > Florent > > > _______________________________________________ > ceph-users mailing list > ceph-users@xxxxxxxxxxxxxx > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |