File on Vmware datastore lost after HP EVA 4400 crash due to power failure

[mchopra]

Our HP EVA 4400 lost power and after resuming power both the controllers had issue with the memory cache and were not starting up.
After replacing the memory cache the controllers are up and showing healthy.
But there was a warning that as the cache was lost then some data may be lost.

The LUNs presented to Vmware host seem to have a problem where some VMs won't power up and give error "File not found".
Currently I only know of 2 VMs with this issue and the vmx file for both was pointing to a snapshot disk (VM1-00001.vmdk & VM2-00002.vmdk).
The -00001.vmdk and -00002.vmdk are missing from the datastores and thus the error "File not found".

Is there a way to retrieve the -00001.vmdk and -00002.vmdk to restore the state before the SAN crash due to power outage?

[Yeskay]

Hi mchopra,

As per the description above, I see that both the controller went down and vdisks are in cache data lost state after the controllers are up. As the data on the controller cache is lost, the vdisks are in cache data lost state. With this state hosts cannot access the Vdisks.

In CV EVA, please make sure that the status of controller and Disk group(DG) are online. If the DG status is good, follow the steps below and clear the cache data lost state on the vdisk.

1. From CV-EVA, on left pane, click on the +sign next to Virtual Disks.
2. Select the vdisk that is in Cache data lost state. The vdisks should be with a red x mark.
3. Click on the option “use vdisk as is”.
4. This option will discard whatever data was there in the controller cache when the EVA went down and get the vdisk online with the data on the disks.
5. Perform this steps on all vdisks that are in Cache Data Lost state and the vdisk should be ready to access by hosts.
6. Then manually check the integrity of the data held on the vdisks from the host.