VMware Data Recovery terminated unexpectedly...

[glowle]

VMware Data Recovery 1.2.1.1616 on a vSphere 4.1.0.381591 host

My backup jobs starts, snapshot machine and starts to backup data, but then at random points the VDR appliance fails with:

Task terminated unexpectedly, possibly due to a power failure or system crash.

Then VDR restarts.

This only occurs when attempting to run a job, otherwise the appliance is rock solid and I remain connected to the appliance in my vCenter.

The VDR appliance is backing up to a datastore on a separate iSCSI LUN presented from a ReadyNAS RN3200.

Many thanks,

Gavin.

[DSTAVERT]

Have a look through the vmware*.log files in the vDR datastore folder to see if there is anything relevant. Check through the ESXi logs. Check your ReadyNAS logs. This would be a somewhat rare issue and you will just need to dig. You might want to file an SR with VMware.

[DSTAVERT]

You can also have a look at the log files directly on the vDR appliance in /var/log/

[Marcj1]

This happens on mine from time to time.  Reformatting the destinations seems to clear it up.

[glowle]

Thank you. I will trawl the logs.

We have support, I may well open an SR if I cannot find anything revealing.

Cheers,

Gavin.

[DSTAVERT]

Since this is a rather important issue you might just initiate the SR.

[Paul_Brannon]

Gavin,

I run into the same problem in the same situation using VDR 1.2.1.1616 on ESXi 4.1.0 348481.

I'm using an Iomega ix4-200r as an NFS server for my VDR destination.  I can't correlate any error on it with the Data Recovery service terminating.  And my /var/vmware/datarecovery logs have rotated during the weekend so the logs at the time of the error are lost.

I'll run the backup again.

Did you find the problem?

thanks,

paul

[glowle]

Hi Paul,

My problems ultimately related to i/o perfomance on the Netgear platform. It had a couple of bonded GB interfaces, but the platform itself was busy with a VMFS partition, NFS and some CIFS. Often the VDR jobs would run fine, but running a scheduled program, which would instigate some concurrent jobs just saturated the box and VDR would fall over. I discovered this without raising the SR in the end.

As a result, I have given my VDR a dedicated LUN from my EqualLogic SAN (not massively cost effective, but cheaper than the headache of not having a vSphere backup!), presented as a VMFS (not RDM) LUN to my VDR appliance.

All running very smoothly now, and with the dedupe, usage not nearly as high as one would expect.

Good luck with your issue Paul.

Best regards,

Gavin.