VM's not powering ON during Test run

[shantanu27]

When I try to do a Test run using  SRM 5.1, the VM's at recovery site doesn't power ON. Also when I try to manually start the VM's it fails to start.

Below are the error messages which pops out when I check the SRM report.

Error #1 Error - "Cannot open the disk '/vmfs/volumes/54607689-9090149b-ac9f-9cb6549655f0/<VMname>.vmdk' or one of the snapshot disks it depends on. " reason.msg

Error #2 Error - reason.msg

Error #3 Error - The object or item referred to could not be found. The object or item referred to could not be found

To add to the confusion there are few VM's(5 out of 26) which power ON without a hitch.

Has anyone come across such issue?

I have approached VMware for this, and according to them Error #1 is because some snapshots on the VM, but when I checked there weren't any.

For Error #2 there is KB from VMware which says the VM's wont power ON as there is some bug related to SRM 5.1, but on my 1st trial the tests ran successfully.(VMware KB: Powering on a virtual machine after running the Recovery Plan in VMware vCenter Site Reco...)

PS:- SRM is using Array based replication

Storage is VNX 5700

vCenter & ESXi hosts are 5.1

[vNEX]

Hello,

Ad1)

Error #1 Error - "Cannot open the disk '/vmfs/volumes/54607689-9090149b-ac9f-9cb6549655f0/<VMname>.vmdk' or one of the snapshot disks it depends on. " reason.msg

I have approached VMware for this, and according to them Error #1 is because some snapshots on the VM, but when I checked there weren't any.

Try to follow this KB article: VMware KB:Committing snapshots when there are no snapshot entries in the Snapshot Manager 

Ad2)

Error #3 Error - The object or item referred to could not be found. The object or item referred to could not be found

This could happen due to datastore discovery timeouts during the recovery phase ... where SRA thinks the operation is complete but the array is not ready yet.

To avoid this you can extend the interval SRM wait for rescan to complete. (SRM 5.1 default is 0)

On  SRM Sites right-click the desired site and select Advanced Settings on storageProvider modify key "storageProvider.hostRescanTimeoutSec"

Set it somwhere between 20-180 sec

For similar issue exists specific KB below which talking about known bug and directs for HBA rescan as workaround ...

VMware KB: Powering on a virtual machine fails with the error: The object or item referred to could ...

And the last comment for KB you have mentioned "Site Recovery Manager fails with the error: Error - reason.msg (2041197)" to eliminate the possibility that you have not enough resources on the

Revovery Site you can change SRM setting to limiting the number of power on/power off operations that SRM concurrently performs, for more info see:

Modify Settings to Run Large Site Recovery Manager Environments

Try to lower the number i.e. to 20 or less...

And of course investigating in heart of SRM log files could reveal much more information.

[shantanu27]

Thanks for your reply, after deep troubleshooting it was found that the array replication didn't occur as required. Once it was synced manually, things worked like a charm.