VMware Cloud Community
NickAC
Contributor
Contributor

After failing over and Failing back LUN has been renamed and also left on DR side as inactive

Hi All

We have a primary and DR site, bith sites have 5 ESXi 5.0 hosts and a separate Vcentre. We are running SRM 5.0

Storage is EMC VNX 5500, we are replicating asynchronously  with snap mirror.

We done a full SRM testing using a test LUN / Protection group at the weekend as we are planning a full DR test shortly, we chose the full test but slected the planned option (we did not want to do the isolated test where it does a snapshot and ring fences it)

.

Everything went fine, we failed the VM's over to our DR site, tested then after 20 minutes we re-protected and failed them back to our primary site.

2 Questions I have are

A, it took 8 minutes to fail them over but 25 minutes to fail them back, seemed to spend a fair amount of time on stage “Preparing the protected VM’s for migration” , what's actually happening under the hood at this stage?

B. After everything has been completed I looked a Vcentre and I see that the test LUN which was originally called TESTSRMLUN01 (original I know) however it is now called "snap-433ad3a8-TESTSRMLUN01", also on our DR side there is a LUN of the same name but status inactive, why has it renamed the Datastore and also why would the datastore still be showing on our DR side Vcentre, I would have thought as part of the reprotect and failback we did the LUN would have been de presented to DR side

Thanks

Nick


Reply
0 Kudos
2 Replies
NickAC
Contributor
Contributor

Hi All

Logged a call with VMware

If anyone is interested there was 2 config items that they advised we change

storageprovider.FixRecoveredDatastores - tick to enable

StorageProvider.HostRescanRepeatCnt - change from 1 to 3

Going to try another failover this weekend and see what happens

Reply
0 Kudos
frcouture
Enthusiast
Enthusiast

Hello NickAC,

We had the same issue over here.  We have the same infrastructure except for the SAN, we use NetApp filers.

Before the reprotect I renamed the datastores to the name it was supposed to be and now the VM is in an inaccessible state at the source site.  I found that the VM files and folder are still on the datastore at the remote site and the VM icon in vCenter is a protected VM. There is actually nothing in the source datastore so SRM and NetApp replication complain about the fact that there is nothing in the source datastore.

Now I'm wondering how can I get the VM back at the source in order to complete the Reprotect.

Is there anyone who ever seen a situation like this?

Regards,

Frank

Reply
0 Kudos