Environment
Running vSphere 4.0(no updates)
HP EVA4400 providing all storage to all hosts
20 production vms
10 pre-prod vms
Storage is all configured as MRU, RR not in use as we running a number of vms as MSCS.
Recently we had EVA controller A fail due to a cache battery, as expected all vSphere Paths failed over to the second controller, B . So now all storage traffic is going via the second controller(B), how do we fail back some of the virtual servers back to controller 1(A), there seems to be no way in either vsphere or the eva.
Currently all pre-prod vms are off and it is reported that there are no performance issues. When pre-prod was brought up the servers(prod and pre-prod) were going via the same controller(b) , though I was away and no actually data was collected to baseline this, but they reported performance issues.,
When you look at the EVA the vdisks are all set to either Path A-Failover/failback or Path B-Failover/failback, but all vdisks have the Managing Controller as B.
If for example I mark Controller B paths inactive in the VIClient for one lun then the path does fail over to the Controller A path but as soon as I make Controller B paths active again the active I/O path is marked as the one to Controller B.
The same if I change the path policy to fixed, mark Controller B paths as down and make Controller A the preferred path, then make Controller B paths active and the path policy MRU the path defaults to Controller B as the active one.
Reading this blog;[http://frankdenneman.nl/2009/02/hp-continuous-access-and-the-use-of-lun-balancing-scripts/] though it refers to ESX3.5 it says I
"If the EVA array detects, in a period of 60 minutes, that at least 66% of the total read request to a LUN are proxy reads, ownership is transitioned to the non-owning proxy controller and making it the owning controller"
So is this saying I just need to leave the environment for a while and it will work which is the right controller for each vdisk?
Thanks
Nigel
