2010-09-15 09:56 AM
Can you share your experiences regarding the impact of Storage VMotion operation on a NetApp filer?
- FAS2050, ONTAP 7.3.4
- vSphere 4.1
- both source and destination datastores are NFS mounts (with plenty of free space) and sit in the same aggregate (with plenty of free space)
When I am moving a single 17GB VM the filer is literally maxed out and I mean it - CPU hits 100% & CP time is 50% to 100%! All operation takes about 20 minutes & there is definitely nothing else going on on the filer - after vMotion completes CPU utilisation immediately comes back to 0-1%.
I have checked fragmentation of a source volume & reallocate measure output says 2, so this is not the culprit I reckon.
Can anything be improved or is it just how Storage VMotion works?
2010-09-16 08:49 AM
Dear Radek ;
i highly recommened to switch your nfs dataStore to Iscsi san LUN
iam using FAS 2020 and i never seen storage vmotion problem before even if with large VMs ...
keep me intouch with you...