BE 2012 backup of vmware hosts causes momentary loss of communication

[Jordon]

Since SP1 my be 2012 is causing lag on my vmware hosts of some kind that causes a moment of network loss of communication.  I have users that say they get disconnectected from the vmware guest or my network monitoring reports that the host went down for a minute.

The loss of communication coincides with the backup job.

my environment is 3 dell PE 1950 III vmware hosts running vshpere v5.0 on 1gbps links and all vm guests running off local disks.

be 2012 is running sp1a and latest hotfixes and is backing up vmware host over 1gbps link.

when I first started using be 2012 on the same vmware hosts I didn't have this issue.

[CraigV]

Hi Jordon,

2 suggestions here:

1. Run LiveUpdate again and see if any additional patches are outstanding. Download them and install them on the media server, and then push-install them to any remote servers that the RAWS agent is installed on.

2. Check the Known Issues section on the link below and see if your issue is listed there:

https://www-secure.symantec.com/connect/backup-and-recovery/issues

If it is, give it the thumb's up. If you don't get this resolved, post back here with an update, and if you have support, open a case with SYmantec...

Thanks!

[Jordon]

media server fully updated already.

haven't seen any known issues yet that are similar.

One thing I discovered today is that the loss of connectivity directly coincided with the consolidation of snapshots.

AND,

• the VMs experiencing the problem had 1 or more old snapshots created by backupexec that weren't deleted.  
• The VMs also do not properly show backup history when the esx host is backed up.  I've redployed the vmware agents and raws agents from scratch multiple times.

in an attempt to fix the problem I am consolidating all of the old snaptops and will try backing up the vmware host again.

[CraigV]

...what version of vSphere are you running? We had a similar issue with another backup product that was causing snapshot releases to kill a VM totally. We ended up upgrading to vSphere 4.1 (from 4.0).

Thanks!

[Jordon]

I'm running the rtm vsphere v5.0

The snaphots, once fully consildated stayed that way and are backing up without interrupting the vm now.