hi,
> Am 19.06.2018 um 20:40 schrieb Steffen Winther Sørensen <ste...@gmail.com>: > > > >> Den 19. jun. 2018 kl. 16.50 skrev Webert de Souza Lima >> <webert.b...@gmail.com>: >> >> Keep in mind that the mds server is cpu-bound, so during heavy workloads it >> will eat up CPU usage, so the OSD daemons can affect or be affected by the >> MDS daemon. >> But it does work well. We've been running a few clusters with MON, MDS and >> OSDs sharing the same hosts for a couple of years now. > We’re also running mds on the osd hosts but again we only run vm backup on > cephfs. we need it to replace the Synology NFS cluster. Also we want share the files between two datacenter (dark fiber with 10Gbit) (but keep both Ceph clusters separate). So I think, we put three MDS in VMs on our Proxmox VM cluster (5 nodes) with a bit RAM and check, if it is reliable. Its much more complicated to have a shared storage for "legacy" systems, than we thought :-) Also RadosGW with NFS is on Debian Stretch not the first choice, as NFS does not has the features, we need (or my Google was to old). cu denny _______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
