Re: System slow

Probably not a network issue in this case. The MQ jobs being in a Mutex Wait (MTXW) state are a clue that the system was having difficulty doing something. Say you were running 5,000 jobs in a subsystem and you decided to end that subsystem. The system may get stressed, and many of those jobs might end up in a MTXW state while ending. Many people understand that new Job activation requires significant resources, but few realize that deactivation / clean-up takes significant resources too.

That's just an example of what could have happened, which of course may not be what did happen. It's just that MTXW is a good indication of resource contention, often due to a resource constraint, such as memory.

-Nathan



----- Original Message -----
From: "fbocch2595@xxxxxxx" <fbocch2595@xxxxxxx>
To: midrange-l@xxxxxxxxxxxx
Cc:
Sent: Tuesday, May 28, 2013 11:04 AM
Subject: System slow


Hi Folks, for about 10 mins today our i went into slowdown mode with CPU in the high 90% (which is fairly frequently from time to time) and our MQ jobs went into MTXW and some applications were stalled or hung, our legasuite app for one.  No interactive sessions were knocked off or ended and no errors for controllers were shown on qsysopr during this 10 min period.  Qsysopr shows only a few jobs running during the 10 min period and those jobs are rountinely run during the day.   

As the network folks wonder what's wrong with the i and my boss too...I say the issue may be the network and I'm not totally sure about that but this type of issue has happened b4 where the i is slow then all of a sudden we're back to normal. 

Is there anything that I can do to prove that the issue was not on the i? 

Is there anything in the performance reports that could help me figure out what happened? 

What's the strategy that you folks would employ in this situation?

Thanks, Frank