On Thu, 18 Apr 2013, Gandalf Corvotempesta wrote: > 2013/4/18 Mark Nelson <mark.nelson@xxxxxxxxxxx>: > > Well, even with RDMA you probably aren't going to get much more than > > ~3.2GB/s (or at least that's what I saw on our production clusters at my > > last job). There's encoding overhead so you can't get the full 40Gb/s. > > > > Beyond that, it's just another software layer with the associated > > inefficiencies. Frankly I'm kind of amazed that rsockets can supposedly get > > around 3GB/s. That's impressive performance imho. > > What about SDP ? It should be a transparent TCP replacement that will > operative natively on an IB fabric with huge speed improvement and > almost none overhead: > > http://bit.ly/14zQzME - Page 43 (4.3.1) I'm no expert, but I've heard SDP is not likely to be supported/maintained by anyone in the long-term. (Please, anyone, correct me if that is not true!) That said, one user has tested it successfully (with kernel and userland ceph) and it does seem to work.. sage _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |