cancel
Showing results for 
Search instead for 
Did you mean: 

Backup Exec 2014 job hangs with the message "discovering devices"

troy_2
Level 3

Hi,

 

Running Backup Exec 2014 SP2 on a Windows 2008 R2 phyiscal server with a HP Robotic Library.    Upgraded from Backup Exec 2010 R3.  Jobs were running fine a few days after the upgrade, but now, after about 10 days after the upgrade, the backup job hangs always at a specific interval into the backup job (~302GB) with the "Discovering Devices" message.  It will just hang there forever.   

 

I've tried repairing the install, repairing the DB, installing the microsoft medium changer driver, and running tapeinst.exe to re-apply the symantec drivers. power cycling the server and robotic library.... nothing works.  the job always hangs at the 302GB point.

Could this be something else with my robotic library?  After the upgrade, would i need to reset my media sets in any way?

 

Thanks,

Troy

1 ACCEPTED SOLUTION

Accepted Solutions

troy_2
Level 3

We replaced the changer because we believed it was a hardware issue, because most the posts related to this message pointed towards that direction towards resolution.  However, changing the hardware out did not resolve the issue.  We also replaced all the tapes, and that also did not resolve it.

What did resolve this is:

1.  Remove the system state backups as part of the backup set

2.  Re-create all backup jobs from scratch

Now, the caveat to what we did to make this work is that this installation was an upgrade.  It was originally BE 2012 and we upgraded it to BE 2014.  Agents were upgraded on the servers summarily.  There could have been a corruption somewhere when the upgraded occurred, but if there was something, there were no errors blatantly evident

So what we did to resolve this may have been particular to our environment.  

But I just wanted to post to follow up how this was eventually resolved.

View solution in original post

5 REPLIES 5

CraigV
Level 6
Partner    VIP    Accredited

Hi,

 

No, no need to reset any media.

BE 15 is out so why upgrade to BE 2014 SP2? You can stop the BE services and run a diagnostic write check of the library to see if there are any issues with it.

Thanks!

pkh
Moderator
Moderator
   VIP    Certified
If your job is backing up m multiple resources, then separate out the resources into individual jobs, so that you can pinpoint which resource is causing problems. Also you might want to try to back to disk to see whether this makes a difference.

VJware
Level 6
Employee Accredited Certified

Discovering Devices usually relate to an issue with the tape hardware. I would recommend as above to test backups to a disk storage.

Larry_Fine
Level 6
   VIP   

I have never heard of a "discovering devices" message appearing fter the BE services were fully started.  I would check my Windows event logs, especially looking for evidence of BE services (specifically pvlsvr) crashing and/or restarting.

I would also check BE's adamm.log for errors.  Possibly BE is losing and regaining connection to your hardware.

What is your tape library and drives (make, model, interface, etc)?  What HBA are you using?

troy_2
Level 3

We replaced the changer because we believed it was a hardware issue, because most the posts related to this message pointed towards that direction towards resolution.  However, changing the hardware out did not resolve the issue.  We also replaced all the tapes, and that also did not resolve it.

What did resolve this is:

1.  Remove the system state backups as part of the backup set

2.  Re-create all backup jobs from scratch

Now, the caveat to what we did to make this work is that this installation was an upgrade.  It was originally BE 2012 and we upgraded it to BE 2014.  Agents were upgraded on the servers summarily.  There could have been a corruption somewhere when the upgraded occurred, but if there was something, there were no errors blatantly evident

So what we did to resolve this may have been particular to our environment.  

But I just wanted to post to follow up how this was eventually resolved.