Backup Exec 2012 - Backup Jobs Stuck in Queued status randomly

[steven_man]

BE 2012 created 3 backup jobs to backup VM to deduplication storage. 3 Backup jobs are schedule to run at the same time. These backup jobs sometimes can be ran successfully. But these backup jobs randomly stuck in queue status over 10 hours , so that nothing can be backup.

Then I want to cancel these backup jobs but failed to do. The status is still queued, not change to "cancel pending". Finally, I need kill begine.exe in task manager and restart all BE serices / reboot BE server.

BE 2012 has been fully patched (include the lastes hotfix 201596). All VM has been installed the lastest version of remote agent.

I found Symantec KB similar to my case. (http://www.symantec.com/docs/TECH159843) 

The solution:

Using the regedit tool, edit the following key:

1. • Key Name: HKEY_LOCAL__MACHINE\SOFTWARE\Symantec\Backup Exec for Windows\Adamm
   • DWORD: Enable OST Space Reclamation
   • Value: 0
2. Restart Backup Exec services.
3. Run the jobs that need to be run.
4. Change the value of the Enable OST Space Reclamation DWORD to 1 to enable it again.
5. Restart the Backup Exec services.

But this KB is applied for BE 2010. Can I apply the same solution  on BE 2012?

Do anyone know the solution?

I have open Symantec support ticket over 2 weeks. But they still not find the solution.

[Gurvinder]

Do you see any alerts while the job is queued ? Has it always been like this ? It may be that BE encounters some malformed OST media which it tries to recreate and that takes time. You can turn on SGMON from BE Install Path and enable device and media verbose with first chckbox and capture to file at the starting of the job. When the job is queued you can provide the sgmon log to tech support for review

[steven_man]

No alert found when the job is queued. The issue randomly happened. Not always happened.

I wil try to turn on SGMon to monitor it.

[lmosla]

Hi, since deduplication is resource intensive try running 2 instead of 3 jobs concurrently or change the start times to a time when there is less activity. It may be that your resources are getting overloaded at different times and causing the jobs to get queued.

[steven_man]

Imosla, I will try it later! I hope you solution can solve my problem