191/158 erros with large image duplication via SLP

[bpup]

Running NBU 6.5.1 on Windows 2003, 1 master/media, 3 SAN media servers.

Primary copy goes to VTL, duplicated via SLP to tape.

Daily backup jobs (inc diff) run fine, duplicate with no issues.

Cumulative diff (weekly) jobs backup with no issues, but do not dupe successfully. They run for a couple hours, even spanning multiple tapes but then eventually fail. No multiplexing is used. Errors in the log show either 158 (couldn't get a lock on file) or 191. But that is all I can find....what causes the 191/158? NBU docs say to look in the duplication progress log...which log is this?

Even though the SAN media servers are set to use their own storage units for dupe, the master seems to run the process and log the errors. Should the master be an "alternate read server"?

Thanks for any info.

[bpup]

Here is the job detail from one that failed after 3 hours and spanning tape:

27/03/2008 3:44:32 AM - requesting resource LCM_nas3-dlt-tape
27/03/2008 3:44:32 AM - awaiting resource LCM_nas3-dlt-tape - resource is busy, will retry allocation later.
27/03/2008 3:49:25 AM - begin Duplicate
27/03/2008 3:49:25 AM - granted resource LCM_nas3-dlt-tape
27/03/2008 3:49:25 AM - started process RUNCMD (5748)
27/03/2008 3:49:25 AM - begin Duplicate
27/03/2008 3:49:26 AM - requesting resource nas3-dlt-tape
27/03/2008 3:49:26 AM - requesting resource DX0025
27/03/2008 3:49:26 AM - reserving resource DX0025
27/03/2008 3:49:26 AM - reserving resource DX0033
27/03/2008 3:49:28 AM - reserved resource DX0025
27/03/2008 3:49:28 AM - reserved resource DX0033
27/03/2008 3:49:28 AM - granted resource NM0979
27/03/2008 3:49:28 AM - granted resource HP.SDLT600.4
27/03/2008 3:49:28 AM - granted resource nas3-dlt-tape
27/03/2008 3:49:28 AM - granted resource DX0025
27/03/2008 3:49:28 AM - granted resource QUANTUM.DLT7000.9
27/03/2008 3:49:32 AM - started process bptm (16076)
27/03/2008 3:49:40 AM - started process bptm (11012)
27/03/2008 3:49:44 AM - started process bptm (11012)
27/03/2008 3:49:44 AM - mounting DX0025
27/03/2008 3:49:45 AM - started process bptm (16076)
27/03/2008 3:49:45 AM - mounting NM0979
27/03/2008 3:49:53 AM - mounted; mount time: 00:00:09
27/03/2008 3:49:53 AM - positioning DX0025 to file 1
27/03/2008 3:49:53 AM - positioned DX0025; position time: 00:00:00
27/03/2008 3:49:54 AM - begin reading
27/03/2008 4:40:49 AM - current media NM0979 complete, requesting next media Any
27/03/2008 4:42:12 AM - started process bptm (16076)
27/03/2008 4:42:12 AM - mounting NM0930
27/03/2008 4:42:11 AM - granted resource NM0930
27/03/2008 4:42:11 AM - granted resource HP.SDLT600.4
27/03/2008 4:42:11 AM - granted resource nas3-dlt-tape
client process aborted(50)
27/03/2008 7:07:32 AM - Error bpduplicate(pid=5748) Duplicate of backup id nas3_1206136841 failed, termination requested by administrator (150). 
27/03/2008 7:07:32 AM - Error bpduplicate(pid=5748) Status = no images were successfully processed.     
27/03/2008 7:07:33 AM - end Duplicate; elapsed time: 03:18:08

** despite saying that termination was requested by administrator, it was not requested by us. All failed dupe jobs have this error code listed, but exit with 191**

[Andreas_Holmstr]

Did you find any solutions to your problem?

I have a similar problem but I work in a Solaris environment also with VTL to SLP Tape duplication?

Kind Regards

Andreas 

[bpup]

Yes, it is a known bug, ET1193887. It was initially thought that the bug only applied to situations where an image was duplicated to more than one destination, but it also affects having more than one image to be duplicated, essentially making SLP completely useless (weekly fulls will not dupe successfully). The fix was to be out 2nd quarter, which has come and gone. I have 2 customers waiting for this fix.

[Andreas_Holmstr]

Thank you for the answer... not the one I was hoping for though...

:( 

Then I need to sort this out in some other way!!! 

Thanks again

 /andreas

[David_McMullin]

And when I searched the knowledge base site for ET1193887, I found nothing...

AARRGH!

[CRZ]

Hi David,

This error is documented under its parent Etrack 1167945.  (I also know there's no easy way for people to know that.)  Here's the TechNote:

BUG REPORT: Duplications fail after 20 minutes with a NetBackup Status Code 50 (client process aborted), if multiple copies are being made concurrently.

 http://support.veritas.com/docs/297164

It should be fixed in 6.5.2.  If you're still having problems, it could be one of the OTHER SLP bugs that eventually got fixed in 6.5.3.