Jbase 6.1.2.2

[alan...@gmail.com]

Looking for some help regarding a spooler lockup issue.

Aix 7.3  Jbase 6.1.2.2

roughly every 48 to 60 hours, the spooler hangs on various print queues. (It appears to be the highly used queues.) I am trying to work with Rocket support and they seem to feel it is a semaphore locking issue of some sort. Jbase spooler commands become inoperable when this occurs, and in many cases the queue goes to a "KILLED" state. No one has been able to determine what is happening through various traces, (jDLS...).

Has anyone run into this issue - better yet a solution. It is affecting business continuity.

The only way out is to reboot the on-prem server...

Any assistance is greatly appreciated!

-Al

[alan...@gmail.com]

Is anyone on Aix 7.3 / Jbase 6.1.2.2

...or just me..

[K, Waruingi]

Hey Alan!

Have you tried running ipcs -ls on AIX to see the current semaphore limits, such as the maximum number of semaphores available?

Methinks that If limits are too low, it could explain why the spooler hangs, especially under heavy load. You might need to increase these limits by adjusting kernel parameters in system configuration files like /etc/security/limits.

Can you also check if any semaphores are stuck and not being released, which could cause lockups...?

Use ipcs -s to list all semaphores and look for any that seem to be in use for too long or not releasing. This may help show if there's a bug in Jbase or another application causing semaphore issues.

Regards,

George W.

--
--
IMPORTANT: T24/Globus posts are no longer accepted on this forum.
 
To post, send email to jB...@googlegroups.com
To unsubscribe, send email to jBASE-un...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/jBASE?hl=en

---
You received this message because you are subscribed to the Google Groups "jBASE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jbase+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/jbase/13399261-c7d4-43fd-aa8f-4d0b77c4a9e3n%40googlegroups.com.

[Alan Metz]

George,

   Thanks for the reply!  Yeah, a semaphore lock gets "stuck" and does not release which leads to the hung state. In general I do not see any semaphore locks for more than a second,

so I don't think we are reaching any kind of maximum limits. (Even when the spooler does hang, it is only one semaphore lock that is held.)

I am working with Rocket support; however, they have not been able to figure out why that lock doesn't release....

Thanks,

Al

To view this discussion visit https://groups.google.com/d/msgid/jbase/CAOqXi%2BA30mGRCyQZLX3a1zYkQ-kL6E222f%3Dx_0UdniOwK2%3DpxQ%40mail.gmail.com.