NBU Java Admin Console

[Stephen_Lindo]

I run the Java Admin Console from my local machine.  I don't use the Windows Admin Console or jnbSA.  I've noticed that when there is no activity (backups/duplications/restores) if I click on any completed job, it comes up blank.  The console then gives me some sort of timeout error and I have to hit the refresh button.  Has anyone else experienced this?  I've seen it in 6.5.x, 7.0 and now 7.5.0.6.  As long as backups or any other activity is happening Activity Monitor is fine.

[sri_vani]

Hello Stephen,

Pls find the below ref cases.. raising symantec call may resolve your issue by providing reqd fixes..

https://www-secure.symantec.com/connect/forums/nbu-java-console-rrror-no-entity-found-restore-screen

https://www-secure.symantec.com/connect/forums/restore-no-entity-found#comment-8168831

verify  Installpath/veritas/netbackup/logs/bpjava-susvc  logs --we may see any clue

is it happening for any particular client/all clients?

[Stephen_Lindo]

Neither of those cases apply to me.  My issue is if my system is idle and Activity Monitor is open, eventually I get a timeout error when I select any completed job.

[Will_Restore · Accepted solution]

>>The console then gives me some sort of timeout error and I have to hit the refresh button.

Had that happen on numerous occasions with various versions.  I've just learned to hit Refresh if the screen has been idle for any length of time.

[SymTerry]

Hello Stephen,

Just to get some further information, what is the timeout error that you are getting? 

[Stephen_Lindo]

The Activity Monitor failed because of Connection to bpjobd daemon on server "master server" failed with the following error:

Software caused connection abort: recv failed.

For more details, refer to the log files.

I've seen this happen on 6.5.4, 7.0 and 7.5.0.6 at three different companies.  The only constant is Java Admin Console.

[Stephen_Lindo]

As have I, but I would like to know why it does this.  I've left the Activity Monitor before thinking no backups were running when it was actually "not connected".

[rk1074]

Could be that ur bpjobd is goin down at times.

Try this out and share the results?:

1. Edit the /usr/openv/netbackup/bp.conf file on the master server and add the entry VERBOSE = 5

2. Create log directories:
# mkdir /usr/openv/netbackup/logs/bpjobd
# mkdir /usr/openv/logs/vnetd

3. Restart the NetBackup processes to enable logging
# /usr/openv/netbackup/bin/goodies/netbackup stop
# /usr/openv/netbackup/bin/goodies/netbackup start

4. Start the Java GUI and go to the Activity Monitor

When the error appears, check the resulting /usr/openv/netbackup/logs/bpjobd/log.<date> file and /usr/openv/logs/vnetd/log.<date> file that may be created. This will list the error that is occurring. This log can then be examined to determine the cause of the error

http://www.symantec.com/business/support/index?page=content&id=TECH29768

[Stephen_Lindo]

The log in question is brutally long.  I guess I'll just have to live with this "enhanced feature".

[SymTerry]

Hello Stephen,

The logs can be very long. If you have the time for a webex, please PM me and I can get a support case opened for you.

[Deb_Wilmot]

That type of error isn't really that unusual, especially if you leave the java console up for long periods of time.

Note that in 7.5 bpjobd was completely rewritten to provide better performance.  Unfortunately we're still dependent upon may external factors - such as network, system resources, etc.,  I also hope that you have the binary installed on the master server that's listed on the 7.5 Late Breaking News site: http://www.symantec.com/docs/TECH178334

(ET3300058) <<Fix Downloadable>> After upgrading to Netbackup 7.5.0.6, the Activity Monitor displays either UNKNOWN or inconsistent Job Details.
 http://www.symantec.com/docs/TECH210574

In the case where you see the following:

"The Activity Monitor failed because of Connection to bpjobd daemon on server "master server" failed with the following error:

Software caused connection abort: recv failed.

For more details, refer to the log files."

I would guess that you are leaving the console up, maybe over lunch or overnight?

What the error is telling you is that there was some sort of communication error and the connection has aborted.  The logs would probably give us little more detail except for confirmation that suddenly we couldn't read or rewite to the socket.. 

In 7.5 the bpjobd process has a monitor thread open for each activity monitor that updates are being pushed out to.  If the connection is dropped or broken, the activity monitor would display the type of message you're seeing (i.e. connection aborted cause nothing is received - probably the socket is gone).  Other monitors (other Java Consoles) may see no issue - again it just depends where the disconnect occurs. 

Another question comes to mind - does your PC 'sleep' after a certain period of inactivity?  I.e. something that would cause the computer to not be pingable?  I know that many companies have PC's with power save type settings that would cause this sort of behavior.

You may also want to check to see if there are any firewalls with inactivity timers set on them.  That could be another thing.

[Marianne]

The weird thing is that I have used Java Admin Console to manage our own NBU installation over multiple NBU versions as well as numerous customer installations for many years...

I have never seen the issue that you describe.

There was only one customer where the NBU master server was behind a firewall and would 'hang' or give error messages after a period of inactivity. We realized that this was a firewall issue, not NBU and simply closed and re-opened the Java Console.