cancel
Showing results for 
Search instead for 
Did you mean: 

Two issues with NBU 7.0.1 and VSphere 4.1

bpup
Level 4
Partner

Running NBU 7.0.1 with EEB 2105102.6 on Windows 2008 R2 x64. Single master/media/vm backup host.

Backups of VSphere 4.1 virtual machines work fine over iSCSI. During restore however, found two issues that cause major restore headaches.

  1. I created a dedicated datastore for the restore job, as it was just a test restore. The restore interface would not pick up the newly created datastore, just the ones that existed when the restore gui (BAR) was first launched. The only workaround was to point to a different ESX server to force a refresh of the datastores, then select the original ESX server once more to trigger a refresh. If I only had one ESX server I would be out of luck.
  2. Experienced the slow vm restore over SAN transport due to the "lazy zero" and "allocate block" issue. The documented workaround is (for BE anyway) to restore directly to the ESX server, not the VCenter. Problem is, in NBU 7.0.1, there is no way to in the restore GUI to get rid of VCenter, as the VCenter dropdown and ESX drop down boxes are separate. The GUI forces the VCenter dropdown to be populated with my VCenter server. How do I point directly to ESX instead?

 

Any fixes on the horizon?

bpup

9 REPLIES 9

RiaanBadenhorst
Level 6
Partner    VIP    Accredited Certified

Hi,

 

Can you please post the note you found for BE? I couldn't find it.

 

Also, I've seen this on LAN transport, so we're illiminating possible causes i suppose.

 

Have you tried to register your ESX hosts in NetBackup in addition to the vCenter? (for purpose of restoring straight to ESX)

bpup
Level 4
Partner

The Tech Note number for the BE side of it is TECH125455

RiaanBadenhorst
Level 6
Partner    VIP    Accredited Certified

Hi,

 

Might want to keep an eye on this post as well, similar issue. Still no clear solution though.

 

https://www-secure.symantec.com/connect/forums/netbackup-7-vsphere-41-restoring-vmdk-issue

Altimate1
Level 6
Partner Accredited

bpup
Level 4
Partner

The reason I could not select "none" for the VCenter drop down was because I only had credentials entered into NBU for VCenter. I added an additional, separate set of credentials for ESX server and then when I started the restore GUI I could select "none" for VCenter. NBU then queried ESX for the resources, which revealed the next issue: the Resource Pool and Datastore drop downs are blank. Even though both VCenter and ESX can see the datastores, NBU appears unable unless pointed via VCenter.  

Giroevolver
Level 6

I have also seen both of these issues. seems there si some sort of bug here that needs to be addressed as restore speed is exstremely slow with the lazy zero!

 

for the first issue does it not work if you just close BAR and reopen?  

VirtualED
Level 5
Employee Accredited Certified

With the version of NetBackup you are using, all you need to do is configure ESX Restore credential in NetBackup.  At restore time, even if you chose VC as the restore destination.  NetBackup will use Virtual Center to create the VM and it will use ESX to perform the restore.

 

So the missing element is that you need to configure ESX Restore Server credential in NetBackup.

RiaanBadenhorst
Level 6
Partner    VIP    Accredited Certified

Hi,

 

The resolution from the other post I mentioned above is to use NBD trasnfer method. This was the instruction from support and i apparently worked.

Giroevolver
Level 6

Sorry but this is incorrect. I have tested this by selecting the VC and had the ESX restore credentials on one of my hosts and it still causes the lazy zero issue.