"Unable to get disk information" when restoring

[rawtaz]

Hey guys,

In short, I am seeing the message "Unable to get disk information" in the "Datastore" step of the Restore dialog:


Short summary of what I've done to get this warning:
- Backed up VMs from an ESXi host just as usual. Worked fine.
- Changed IP on this ESXi host, and also recreated its datastore.
- In Veeam (the B&R console/application), removed the old ESXi host from Servers, and added the new one (practically the same but with another IP and datastore).
- Tried restoring VMs from the backup, but got the error "Host with id 'some long number' was not found" when selecting which VM to restore (the VM is actually put in the list, but in subsequent views throughout the Restore dialog it doesn't show up, which is a showstopper).
- Removed the backup from under Backups and used Import Backup to get it back.

Now the previous error no longer shows, but I do get the warning mentioned first in this post. If I continue with the restore, Veeam does seem to restore all the disks for the VM (two in this case) - I just can't see information about them, and can't set their new type, in the restore dialog).

Is this expected or unexpected, based on the information above?

I also discovered a workaround for this. If, when in the Datastore step and seeing this error, you go back three steps to the Restore Mode view, and then forward again to the Datastore view, the disks are updated and visible (i.e. normal behavior). My impression is that it's because this time, when the switch from Restore Mode -> Host happens, the correct/new host is already selected for the VM. I base this on the fact that I tried going back only two and one steps, but this had no effect. It's when you go back at least three times, so that the Restore Mode -> Host view change happens, that it "works" again.

It would be nice if someone could confirm this. I do not know if this is a local/isolated issue or can be reproduced in other environments as well. I can reproduce it here though.

I should also mention that this warning is shown regardless of whether I do Import Backup or Rescan Repository in order to get past the first error I mentioned ("Host with id 'some long number' was not found").

Let me know if you have any questions. Thanks.

[foggy]

Looks like a bug. QC will try to reproduce this behavior, I'll keep you posted here.

[rawtaz]

Cool. Any information you need, just let me know.

A clarficiation, albeit obvious: When I first go to the Restore dialog and select the VM, then in the Restore Mode I need to select "Restore to a new location", and then in the Host view the old ESXi host is selected (or rather, the old IP). Here I must change this to the new host in order to get to the subsequent views.

If after doing this I don't proceed to the Resource Pool and Datastore views, but instead directly go back one step to the Restore Mode view, and then move forth to the Datastore view, I get the disks visible as expected. I.e. I don't have to go to the Datastore view before doing the "workaround".

[foggy]

Honestly, I had this kind of issue myself at the beta stage but then they seemed to fix it. Looks like it is back again for some reason..

[spgsit5upport]

Same here, on latest patch3
No matter what OS VM is being restored (which should not matter at all)

Seb

[foggy]

This is confirmed as an issue and will be addressed in one of the future updates.

[lukongodo]

This is confirmed as an issue and will be addressed in one of the future updates.

Hi was anyone able to sort this out. I am also getting the same error.

Ken

[Gostev]

Hi Ken, this topic is pretty old... are you using the latest product version?

[foggy]

Just as an update, this seems to be addressed in v7.

[VladV]

Latest Veeam B&R Patch3 and got the same issue. Thanks rawtaz for your workaround, it worked for me. In my opinion this should have been fixed quickly and before V7.

I got this error when I tried to restore vCenter itself. Because vCenter was down, I had to add in the backup infrastructure the destination host as a standalone. I then proceeded to restore Entire VM and noticed this warning "Unable to get disk information". I continued ignoring it, but after a while I received this error:

Code: Select all

5/27/2013 4:35:59 PM Fail     Restoring disk '[System] VMware vCenter Server Appliance/VMware vCenter Server Appliance.vmdk' (25.0 GB)
5/27/2013 4:37:07 PM Fail     Restore job failed Error: Client error: Failed to open VDDK disk [[System] VMware vCenter Server Appliance/VMware vCenter Server Appliance.vmdk] ( is read-only mode - [false] )
                              Failed to open VMDK.
                              Logon attempt with parameters [VC/ESX: [192.168.1.101];Port: 902;Login: [****];VMX Spec: [moref=52];Snapshot mor: [52-snapshot-1];Transports: [nbd];Read Only: [false]] failed because of the following errors:
                              Cannot process [restoreDiskContent] command. Target descriptor is [vddk://<vddkConnSpec><viConn name="1

Followed your workaround and it worked.

[foggy]

Vlad, glad that the workaround provided by OP worked for you. The fix from our side is scheduled for the next version.

[KeithW]

This is still a problem with Version 7, patch 1. I just tried to restore my vCenter VM and had exact same issue, with the exact same successful workaround.

In my case though the restore failed due to a zlib failure, I suspect related to the fact that the VM tanked on the backup snapshot commit phase, which of course tanked vCenter, and abended the backup job. In any case, I think that is unrelated to the OP issue, but wanted to note that it is still an issue.

[foggy]

Keith, we weren't able to reproduce this on v7 (with any of the patches applied). Could you elaborate on what changes could lead to this behavior (changing host IP, like in the OP's case, or whatever)?

[sideshowtob]

I just had the same experience also when restoring my vCenter. I updated to the latest patch (veeam_backup_7.0.0.764.R2.exe) and still had the same problem. Workaround posted in the thread worked fine.

My vCenter was corrupted and down, so I had to add the recovery host by IP. Could it be anything to do with the fact its vCenter being recovered?

[Gostev]

Most likely, yes. If I remember correctly, the wizard is trying to query the datastore that the virtual disk was backed up from (using the metadata in backup file). And because your vCenter is down, this is impossible.

[foggy]

QC finally was able to reproduce this behavior after changing host IP and datastore name, re-adding it to Veeam B&R console, and trying to restore the VM backed up via the original host IP. It would be great if someone could open a case and provide logs to facilitate the process of addressing the issue. Thanks.