You say “even when the cluster is doing nothing” - Are you seeing those numbers on a completely idle cluster? Even SSDs can go to sleep, as can CPUs (throttle/sleep states), memory gets swapped/paged out, tcp connections die, cache is empty... measuring a completely idle cluster is not always representative. In my cluster I have seen these number go high and low without any measurable performance impact, but in my case it’s somewhat more CPU-bound than you are on a much newer version.
I suggest you measure the performance of a RBD device with fio or such, latencies are much more important than throughput (until throughput becomes the bottleneck). Jan > On 03 Jun 2015, at 14:19, Xu (Simon) Chen <xche...@gmail.com> wrote: > > Hi folks, > > I've always been confused about the apply/commit latency numbers in "ceph osd > perf" output. I only know for sure that when they get too high, performance > is bad. > > My deployments have seen many different versions of ceph. Pre 0.80.7, I've > seen those numbers being pretty high. After upgrading to 0.80.7, all of a > sudden, commit latency of all OSDs drop to 0-1ms, and apply latency remains > pretty low most of the time. > > Now I'm trying hammer in a new cluster, and even when the cluster is doing > nothing, I see commit latency being as high as 20ms, and apply latency being > 200+ms, which seems a bit off to me. > > Any ideas how these numbers changed over versions? > > Thanks. > -Simon > _______________________________________________ > ceph-users mailing list > ceph-users@lists.ceph.com > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com _______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
