On Mon, May 26, 2014 at 10:53 AM, Stefan Priebe - Profihost AG <s.priebe at profihost.ag> wrote: > Hi, > > sorry it was a bit poorly defined. > > I'm talking about thinks like this: > http://lwn.net/Articles/551179/ > > Stefan > Not sure if Ceph can have any advantage of it, as for common Ceph operations looks like they are barely hitting the performance area patch is designated about of, but it would be awesome if you`ll be able to test it :) D(2)CTCP and may be some other congestion control algorithms designed for low-latency high-speed networks can definitely give you a speed bump for spiky workloads. > Am 25.05.2014 11:11, schrieb Andrey Korolyov: >> Hi, >> >> which one you are talking about? -rt patchset has absolutely no >> difference for Ceph, though very specific workload (which I was unable >> to imagine at a time) can benefit of it a little. Windriver variant >> means much more, because it rt`ing virtualized envs - in combination >> with storage nodes you may achieve a lot better deadlines for tasks >> like gaming servers and so on, but I had not tried it. >> >> On Sun, May 25, 2014 at 1:03 PM, Stefan Priebe - Profihost AG >> <s.priebe at profihost.ag> wrote: >>> Hi, >>> >>> has anybody ever tried to use a low latency kernel for ceph? Does it make any differences? >>> >>> Greets, >>> Stefan >>> _______________________________________________ >>> ceph-users mailing list >>> ceph-users at lists.ceph.com >>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |