On Wed, Feb 20, 2013 at 4:21 PM, Jonathan Rudenberg <jonathan@xxxxxxxxxxxx> wrote: > Thanks for the input! > > On Feb 20, 2013, at 2:13 PM, Mark Nelson <mark.nelson@xxxxxxxxxxx> wrote: > >> I'd be much more worried about memory during recovery than CPU assuming that the CPU is specced relatively close to our guidelines. > > How much memory is needed? The current plan for these nodes is 256GB, as the 32GB DIMMs aren't economical. > > Jonathan That's definitely enough. Our rules of thumb are 1GHz and 1GB, although I've heard from one source that large-scale cluster issues can really want more like 1.5GB. 7GB per daemon will provide lots and lots of juicy caching. ;) _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |