On 10/24/2013 03:36 PM, Martin Catudal wrote:
Hi,
Here my scenario :
I will have a small cluster (4 nodes) with 4 (4 TB) OSD's per node.
I will have OS installed on two SSD in raid 1 configuration.
I would never run your journal in RAID-1 on SSDs. It means you'll 'burn'
through them at the same rate, so there is no benefit.
Is one of you have successfully and efficiently a Ceph cluster that is
built with Journal on a separate partition on the OS SSD's?
Not that I know of, I would always seperate it.
I know that it may occur a lot of IO on the Journal SSD and I'm scared
of have my OS suffer from too much IO.
Any background experience?
What I'd suggest:
* One small (~60GB) SSD for your OS
* One SSD per 6 OSDs for journaling
I have never seen one Intel SSD fail. I've been using them since the
X25-M 80GB SSDs and those are still in production without even one
wearing out or failing.
Nowdays I'm so comfortable with SSDs I never use RAID on them when
running a OS on it.
With Ceph I also see machines as 'disposable', so IF a SSD for the OS
fails I don't care, since the rest of the cluster will take over.
This gives you another free slot which you can use for an OSD.
Some chassis from SuperMicro have internal 2.5" bays where you can place
your SSD for the OS so you can use the hot-swap bays for your Journal
and Data disks.
Martin
--
Wido den Hollander
42on B.V.
Phone: +31 (0)20 700 9902
Skype: contact42on
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
