[one-users] Performance with system datastore on NFS

[Daniel Dehennin]

Hello,

I just finalize the migration our 3.8.3 ONE to 4.2 and was forced to
change my plan to put datastore 0 on NFS4 because of slow accesses.

For non persistent images they use local datastore 0, but persistent
ones use NFS (via the symlink).

All my images are qcow2.

I have a dedicated VLAN for storage access, and my nodes are mounting
the datastores as:

    10.255.255.2:/one-datastores on /var/lib/one/datastores type nfs4 (rw,relatime,vers=4,rsize=262144,wsize=262144,namlen=255,hard,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=10.255.255.4,minorversion=0,fsc,local_lock=none,addr=10.255.255.2,_netdev)

I remember seeing a document about different cache scenarios and
performance but I can not remember where.

My setup is a tree nodes ONE:

- one not very powerfull frontend
- two quite powerful nodes to run VMs (core i7 + 16Go RAM)

My storage is what they call “workgroup NAS” with dual gigabit nics
configured in bonding.

I'm wondering about using RAID1+0 instead of the RAID5 to improve disk
access performances and lower CPU usage.

Any hints or idea?

Regards.

-- 
Daniel Dehennin
Récupérer ma clef GPG:
gpg --keyserver pgp.mit.edu --recv-keys 0x7A6FE2DF
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 229 bytes
Desc: not available
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20131120/fc4af0bb/attachment.pgp>