W dniu 15.03.2017 o 09:05, Eneko Lacunza pisze:
Hi Michal,
El 14/03/17 a las 23:45, Michał Chybowski escribió:
I'm going to set up a small cluster (5 nodes with 3 MONs, 2 - 4 HDDs
per node) to test if ceph in such small scale is going to perform
good enough to put it into production environment (or does it perform
well only if there are tens of OSDs, etc.).
Are there any "do's" and "don'ts" in matter of OSD storage type
(bluestore / xfs / ext4 / btrfs), correct
"journal-to-storage-drive-size" ratio and monitor placement in very
limited space (dedicated machines just for MONs are not an option).
You don't tell us what this cluster will be used for. I have several
tiny ceph clusters (3 nodes) in production for some years now, ceph
nodes usually do mon+osd+virtualization.
My bad, it will provide disks for VMs, maybe in future it'll be expanded
a lot to provide backend for object storage.
They perform quite good for their use case (VMs only use heavy I/O
rarely), but I have always built the clusters with SSDs for journals.
I have seen better performance with this setup than some entry-level
EMC disk enclosures; I always thought this was a misconfiguration
problem on the other enclosure provider though! :)
Cheers
Eneko
--
Michał Chybowski
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
