I setup ceph cluster on 28 nodes.
24 nodes for osds...Each storage node has 16 drives. raid0 on 4
drives. therefore i have 4 osds daemon on each node. each osd daemon
is allocated a raid volume. so total of 96 osds daemon in the entire
cluster.
3 nodes for mon ...
1 node for mds ...
I mounted a share with CEPH-FUSE and im generating load to the cluster
through the mount point.
Network bandwidth is fairly
mon nodes - 20 gb to switch
osd nodes - 1 gb to switch
mds node - 20 gb to switch
Questions:
1. How to optimize the network for better performance
2. Is 1 mds good for this cluster ?
3. How check for replication and what replication factors should be used?
4. What the debug logging configuration should look like?
All the aim is make ceph-fuse give its best in write,read and delete
operations... Pls kindly guide me.
Thanks
---------- Forwarded message ----------
Hi,
Pls can somebody help interprete what this configuration signify.
[global]
debug ms = 0
[mon]
debug mon = 20
debug paxos = 20
debug auth = 20
[osd]
debug osd = 20
debug filestore = 20
debug journal = 20
debug monc = 20
[mds]
debug mds = 20
debug mds balancer = 20
debug mds log = 20
debug mds migrator = 20
I used the same configuration and my boot disk got full
ceph-mds.a.log = 50299281408
ceph-mon.a.log = 173858816
Moreover the results of my test using ceph-fuse to mount a share and
exportiing the share with samba isn't so good... How can i turn off
some of the logs to optimize ceph performance in a write,read and
delete operations. I will publish the result as soon as i can get the
cluster more optimized.
regards.
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
