VMware Cloud Community
markokobal
Enthusiast
Enthusiast

VDR (daterecovery process) suddenly started repeatedly crashing

Hi,

Suddenly all of our Data Recovery appliances started repeatedly crashing. The appliance VM itself is running OK, but the datarecovery service is crashing soon after the service starts. So all of the appliances are actually unusable :(.

/var/log/messages:

watchdog-datarecovery: '/usr/sbin/datarecovery' exited after 368 seconds

watchdog-datarecovery: Executing '/usr/sbin/datarecovery'

datarecovery: Starting VMware Data Recovery, version 2.0.2.3440

watchdog-datarecovery: '/usr/sbin/datarecovery' exited after 369 seconds

watchdog-datarecovery: Executing '/usr/sbin/datarecovery'

datarecovery: Starting VMware Data Recovery, version 2.0.2.3440

watchdog-datarecovery: '/usr/sbin/datarecovery' exited after 369 seconds

watchdog-datarecovery: Executing '/usr/sbin/datarecovery'

datarecovery: Starting VMware Data Recovery, version 2.0.2.3440

watchdog-datarecovery: '/usr/sbin/datarecovery' exited after 369 seconds

watchdog-datarecovery: Executing '/usr/sbin/datarecovery'

datarecovery: Starting VMware Data Recovery, version 2.0.2.3440

watchdog-datarecovery: '/usr/sbin/datarecovery' exited after 368 seconds

watchdog-datarecovery: Executing '/usr/sbin/datarecovery'

datarecovery: Starting VMware Data Recovery, version 2.0.2.3440

watchdog-datarecovery: '/usr/sbin/datarecovery' exited after 368 seconds

vDR crashes.png

There is no obvious reason for crashes, tried to migrate VDR to different hosts, tried to increase vCPU and vRAM but it doesn't help. No other errors in logs (neither in the vCenter logs).

We were running 2.0.1, we've upgraded to 2.0.2 but still with the same issue. If I do a reboot of VDR it works for some time (non-deterministic - sometimes a couple of minutes, sometimes hours), that starts crashing.

One thing that I noticed after VDR was working (after reboot) is that all of the jobs lost the link to the selected Virtual Machines; so I have to re-configure the jobs and select the desired VMs once again (all other settings - destination, schedule, etc. was still OK within the jobs).

Has somebody else experienced this situation?

BTW: I'm running vSphere 5.1u1, where VDR is no more supported, so VMware support can't help me (and I don't want to migrate to VDP for obvious reasons...) ... I was really happy with VDR and now this show-stopper is really making me crazy ...

-- Kind regards, Marko. VCP5
Reply
0 Kudos
2 Replies
aliensalien
Contributor
Contributor

Hello,

I experienced the same problem as you before and I've got a solution from VMware support.

The fix requires to replace the libchunkDedupe.so file in the VDR appliance and I can't post that file in here. 

So please leave me your email address for details.

Reply
0 Kudos
markokobal
Enthusiast
Enthusiast

Hi,

Interesting, I did as well file a SR on this topic, but VMware said "we have no solution for you" ... anyway, meanwhile we've already switched to VMware Data Protection and we are not using VDR anymore... Thanks anyway!

-- Kind regards, Marko. VCP5
Reply
0 Kudos