I have two vCenter 4.1.0 servers, one at the protected site and one at the recovery site. I have SRM 4.1.2 installed at each site on seperate servers. vCenter and SRM are installed on W2K8 R2 VMs. Each site has a Clariion CX4-120 with MirrorView enabled and the MirrorView SRA v1.4.0.16 installed on each SRM server. Each site uses a Solutions Enabler virtual appliance 7.4. I use Unisphere to configure the Clariions. We use asynchronous mirroring.
Here are my steps taken:
Created LUN on protected site. Configured LUN for correct storage group. Created snapshot called VMWARE_SRM_SNAP and added to correct storage group. Created asynchronous remote mirror and added secondary image LUN to mirror. Created snapshot at recovery site from remote mirror called VMWARE_SRM_SNAP and added to correct storage group.
Added a new datastore on protected site vCenter. Created a new W2K8 VM on datastore. Rescanned hosts on LUNs at protected site in Unisphere. Created new protection group at protected site for new LUN which protected the new W2K8 VM. Created new recovery plan for this group.
I can run a successful test but running an actual failover fails during "recover normal priority virtual machines" with the error Failed to recover datastore. I dig through the vmware-dr.log and find errors such as "Error Details: Failure during Clariion Interface Layer's attempt to obtain data from Clariion array" and "failover exited with exit code 0". My research has lead me to this KB article http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1020941 which talks about issus with asynchronous mirrors. I added the message step as indicated and the failover actually worked if I manually told the mirror to synchronize. A month later, now that I am preparing to go live and continue our data recovery plan, I cannot get the VM to failover at all anymore. I have reversed all my steps and keep getting the same error. In order for me to start over and try again, I have to fully sync the mirror and promote the mirror back to the original primary at the protected site which takes about 5 hours. EMC has been no help so I have come here for any assistance. I have attached the dr.log for review. Thank you for any future assistance with this!