On Thu, May 28, 2015 at 1:04 AM, Kenneth Waegeman <kenneth.waege...@ugent.be> wrote: > > > On 05/27/2015 10:30 PM, Gregory Farnum wrote: >> >> On Wed, May 27, 2015 at 6:49 AM, Kenneth Waegeman >> <kenneth.waege...@ugent.be> wrote: >>> >>> We are also running a full backup sync to cephfs, using multiple >>> distributed >>> rsync streams (with zkrsync), and also ran in this issue today on Hammer >>> 0.94.1 . >>> After setting the beacon higer, and eventually clearing the journal, it >>> stabilized again. >>> >>> We were using ceph-fuse to mount the cephfs, not the ceph kernel client. >> >> >> What's your MDS cache size set to? > > I did set it to 1000000 before (we have 64G of ram for the mds) trying to > get rid of the 'Client .. failing to respond to cache pressure' messages
Oh, that's definitely enough if one client is eating it all up to run into this, without that patch I referenced. :) -Greg > Did you have any warnings in the >> >> ceph log about clients not releasing caps? > > Unfortunately lost the logs of before it happened.. But nothing in the new > logs about that, I will follow this up > >> >> I think you could hit this in ceph-fuse as well on hammer, although we >> just merged in a fix: https://github.com/ceph/ceph/pull/4653 >> -Greg >> > _______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
