corrupt data in RecoverPoint

Question

We have some data at an external facility that we are trying to bring back to our facility.  The source LUNs are from VNX arrays and the targets are VNX5600's.  We've placed RecoverPoint (4.02) out at the external facility and in our datacenter.  What is being done is that the source LUNs are being cloned, we fracture the clones, and then use the clones for the source RecoverPoint LUNs.  We've successfully replicated those LUNs to our facility.  The LUNs are VMware RDM's.  The external facility is using vSphere 4.1 and we are on vSphere 5.5.  To test the validity of the data we do a test copy of the copy in RecoverPoint and then mount the LUN to a virtual machine and the data ends up being corrupt.  We've noticed some files missing from the LUNs and if you run a chkdsk (read only mode) it reports a bunch of corruption.  To validate that the data is valid on the source side we have taken the clones LUNs and mounted them to a virtual machine in there facility and we see no corruption.  We've also take the source RecoverPoint copy and have done a test copy with that and mounted it to a virtual machine and see no corruption.  Does anybody have any idea where the corruption could be coming from or have experience with this particular issues.  We've opened cases with all the vendors involved but haven't gotten anywhere yet.

rgcda · Answer

The target ESXi servers are 5.5.  So are you saying the drive should be resignatured?  They are completely separate vCenter environments and sans so there shouldn't be any issues with a LUN being detected as a snapshot of another one.

rgcda · Answer

when I run esxcli storage vmfs snapshot list on one of the ESXi servers it doesn't indicate any of the LUNs are snapshots.  Where is this EnableResignature setting in ESXi 5.5.  I'm not seeing it under Advanced Settings > LVM?

forshr · Answer

It's a case of maintaining the LVM signature. Can you check on the target ESX servers that the advanced EnableResignature parameter is set to 1.

forshr · Answer

Can you check that ESX resignaturing is turned on the target ESX 4.1 servers.

forshr · Answer

For 5.5 you need to navigate to Configuration->Storage->LUN and then Keep the existing signature under Mount Options.

rgcda · Answer

These are RDM's so I don't have the same options as adding a traditional vmfs volume.

forshr · Answer

Yes, sorry you did state this. So this is what I would do/say on this issue.

As long as the clones are fractured the LUNs are standard LUNs as far as RP is concerned. The process I would follow would be this:

1. Fracture the clones.

2. Then create the CG ensuring the host_os policy for each copy is set to VMware ESX Windows and check that all LUNs are attached to the respective splitters.

3. Then start the transfer.

4. Once the CG reaches an Active state, run the start_integrity_check CLI command and then monitor it with the show_integrity_check CLI command. This should identify any data block difference between the source and remote copies.

5. If there are no issue initiate image access and test the integrity of the data.

RecoverPoint

Was this post helpful?