Networker 7.6.1 group hangs for hours

Question

This group has just stopped for no apparant reason. It does not say it is waiting for media, it just is doing nothing after completing 60% of the backups. Its status is 'running.'  The next up client says some sessions are active (they are not doing anything) and others are queued. This job is only for SQL backups and worked fine yesterday and generally in the past. No new backups were added to any of the clients. There are plenty of tapes, even new tapes, that can be used. It does not say it is waiting for media. Both tape drives are empty. There are tapes available for the bootstrap index. All the prior clients completed successfully and there is only one client left in this group. The SQL server has no log messages pertaining to this job, except to note that yesterday's backups worked fine. There is nothing in the server event log other than successful messages from other jobs. In the past I could restart the job and it would work fine. Is there any way to make these more reliable? This kind of hang seems to happen more often with SQL groups. The client SQL version is 5.2.1.299.

david_knapp · Answer

When put into another group, it just sits there in queued status and does nothing.

Bebo2k · Answer

Hi Davek,

Is it an inactivity timeout issue ? Did you tried to change the saveset and check the backup if it works fine with smaller one ?

Would you please try to run backup for that client with another saveset , and check what will be the results, then run that group with that specific client in verbosity mode with the old saveset, You can run the following command from the backup server command prompt:

savegrp -vvvvv -l full -c clientname -G groupname

Please provide us with the output of that command in a text file

Also please provide the xbsa.messages and the daemon log.

Thanks,

Ahmed Bahaa

CarlosRojas · Answer

Hello Davek,

This looks to me as if you have reached the maximum connection number (parallelism) from NW server to client.

To solve this issue I would suggest you to:

1.- Ensure there is no nsrsqlsv.exe or save.exe running on the SQL client.

2.- Stop NetWorker services on the SQL client.

3.- Rename/delete /nsr/tmp folder

4.- Stop any group running on NetWorker server (or wait for all groups to complete)

5.- Rename/delete /nsr/tmp and /nsr/res/jobsdb folders in NetWorker server

6.- Start up NetWorker services on NetWorker server

7.- Start up NetWorker services on SQL client.

8.- Run a new backup

Alternatively, and depending on the parallelism set for the server, you can increase the parallelism for that client to, let's say 100, this will enable the server to attempt new connections in case you reached already the maximum.

Some times there are some sessions hanging, what prevents NetWorker server to create a new connection or request any further session. Following the step will clean out any current job and will get rid of any hanged session.

Thank you.

Carlos.

david_knapp · Answer

After going through the deletions/renames/etc, it worked for one day, then continued its hanging. The client was set to a parallelism of 30. The job opened 10 instances of nsrsqlsv (same number of databases to be backed up), none of which seem to do anything. I have a bunch of other SQL jobs that backup just fine. The clients are all set the same as this one. The messages log just says it is waiting for 10 jobs to complete. The SQL server is version 2008. When I abort the group with NMC, none of the nsrsqlsv processes stop, I have to kill them manually on the SQL server end. On prior executions of this I see the following in the messages log - Apr 19 10:51:47 servername: * servernamebu:MSSQL$MSSQL2008:Master 37692 1334839127 5 0 0 4052 6168 0 servername.com (pid6168) 56 CreateInstance failed on Connection object in %s line %d 2 23 13 nsrsqlado.cpp 1 3 823.

So, it seemed to be something in the connection, or lack of, after the nsrsqlsv process starts. Next, I tried to uninstall and reinstall the SQL client. The MSInstaller would error off every time. Finally, I rebooted the server, re-installed the SQL client successfully, and was able to make desired backups twice. So, looks more like it was some sort of Windows OS problem after all. Will update later.

sssun · Answer

Its Simple, Just verify whether you have your backup group CHECKED in the media>media Pool>select media pool you have specified and check the group name which has backup issue. Thanks SAS

NetWorker

Networker 7.6.1 group hangs for hours

Was this post helpful?