Instant Recovery migrate to production keeps failing

[MVanek]

Hi,
I am recently having problems within my lab environment. I havent been able to find similar problem here on forum, or is just my misunderstanding.
Just trying to migrate some VMs from Hyper-V to VMWARE.

Its only a lab env so I have veeam server/repo as VM on ESXi - the same ESXi is used as target for Instant recover and Quick Migration. I know its not good but Iam just testing it.

I have backed up one test Windows machine from Hyper-V with no problem (about 100GB of data).
When I do Instant Recovery, it works good without any issue. VM is up and running ( FYI redirect write cache is directed to same ESXi datastore where I want to perform quick migration)

I do use cold migration with powered off VM.

Bwhen I try migrate VM to production it fails with
Copying snapshot files Error: Failed to open VDDK disk....
I might copy some info from log, but Iam not sure if I can do it here on forum. There are VDDK related errors like: VDDK error: 1 (Unknown error). Value: 0x0000000000000001
and more.

Thanks for your tips.
Regards
Martin

[PetrM]

Hi Martin,

It's a complicated problem and likely our support team will need to examine your infrastructure, analyze debug logs and probably involve VMware support. Usually, an attempt to open a disk fails with VDDK error 1 due to an issue with a datastore on which virtual disks reside. So the best action plan is to let our engineers work on it. Also, please don't forget to share a support case ID as requested, all posts about technical issues without case ID will be eventually deleted by the forum moderator.

In the meantime, I'd suggest trying another target datastore for migration, I guess it's related to a disk that belongs to the target VM, not the one which is running in IR.

Thanks!

[MVanek]

Hi Petr,
thanks for your reply.
I have a little but important update. Its gonna seem funny or most likely stupid )
I have restarted ESXi host and problem is solved and everything worked like a charm now...
Reason that I think could be the culprit is that I have built this lab in quick manner - I have installed new ESXi, put a license on it and yesterday tonight I just realized that this ESXi hypervisor never been restarted after the initial installation.

The problem was probably related to VDDK API and this is a service which from what I know doesnt work if you have free license and as I have licensed this host after the clean install and never restarted it then the API could not work as well as it should.

Anyway, sorry for alarming the forum and also I would like to thank for your response.
Regards
Martin

[MVanek]

Hello,
I need to refresh this topic, since I have struggled this issue again and I have found out what could be the culprit of this problem.
As I mentioned earlier - I have rebooted ESXi and the problem gone away.

The reason was, that on that same host there was vcenter instance running, when I rebooted it I had to turn this vcsa off.
Since this is a lab I didnt power on this vcenter vm again imediatelly after I rebooted that ESXi.
So with powered off vcenter Quick migration of Instant Recovered VM worked with no problem.
But when is vcenter powered on it fails with problem above.

Just to make things clear - I have VEEAM server connected directly on ESXi which is managed by vcenter (okay I know its not good, but its a lab for destroying things )

SO:

When I do Quick Migration within single ESXi, which is managed by vcenter (but Veeam connected directly on ESXi), when Veeam start "touching" on snapshot files (redirected cache of this VM) for quick migrating it to production, vCenter starts to notify, that "VM Guest OS crashes" and after that it says ""vSphere HA virtual machine failed to failover" and according to this the Quick Migration Job fails with Copying snapshot files Error: Failed to open VDDK disk.... and blah blah.

FYI there is NOT any cluster or HA enabled within this vcenter. Its only vcenter with connected ESXi. Nothing else.
Can you please clarify me, that this is normal behavior? Will there be any difference if I had redirected Veeam to vCenter instead of standalone ESXi?
I havent had time to test it, so when I do, I let you know.

I hope I have described this understandable way
Regards
Martin

[Fab G.]

Hello,

It happen to me in P2V operation.

Case :
1 HP Gen6 server, 2008R2 with discontinued apps
2 HPE Gen10+, ESX 8.0u1 with vCenter (no HA)
1 HPE Gen10, Windows Server 2022 with Veeam B&R 12

I made a backup with Veeam Agent 6 on old server managed by Veeam B&R, start Instant recovery (all is fine here), start migration to production then at the end :

29/12/2023 12:35:25 :: Copying snapshot files Error: Failed to open VDDK disk [[xxxxxxx] xxx.xxx.xxx.xxx/disk0.vmdk] ( is read-only mode - [false] )
Logon attempt with parameters [VC/ESX: [xxx.xxx.xxx.xxx];Port: 902;Login: [xxxxxxxx];VMX Spec: [moref=21];Snapshot mor: [21-snapshot-1];Transports: [nbd];Read Only: [false]] failed because of the following errors:
Failed to open disk for write.
Failed to download disk 'vddkConnSpec>'.
Agent failed to process method {DataTransfer.SyncDisk}.

Failed to process VM xxx.xxx.xxx.xxx at Friday, December 29, 2023 12:37:12 Error: Failed to open VDDK disk [[xxxxxxxx] xxx.xxx.xxx.xxx/disk0.vmdk] ( is read-only mode - [false] )
Logon attempt with parameters [VC/ESX: [xxx.xxx.xxx.xxx];Port: 902;Login: [xxxxxx];VMX Spec: [moref=21];Snapshot mor: [21-snapshot-1];Transports: [nbd];Read Only: [false]] failed because of the following errors:
Failed to open disk for write.
Failed to download disk 'vddkConnSpec>'.
Agent failed to process method {DataTransfer.SyncDisk}.

Workaround was disconnecting ESX from vCenter, powerdown vCenter, then retry.
I'm not sure that the exact same issue here but if it can help...

[Bizaro]

Had the same issue. Disconnecting the ESXi host from vCenter solves the problem.