Fred Lee
Member
We had an issue where we restarted the kernels and they would not end and they were in MTXW status so we had to end the job abnormally. Did anyone have similar experience and have any luck in resolving this.
PMatino
Member
Fred whats version and tools release are you running?
We had this type of issue it was caused by the way that operations was ending and starting the services.
In short what was happening was the services were started without issueing the CLRIPC command the night before when ending services. The solution I put in place was to do the following. I would had operations move the CLRIPC command to be kicked off before the STRNET command gets issued during startup.
This is all done for us using BMC scheduler. Make sure your shutdown and start up procedures are happening in the proper sequence.
Another reason this could be happening is that when the services are being ended you have a kernel running with outstanding requests eating up CPU and basically doing nothing. A kernel like that will end fine with an end immediate command issued. Just make sure there are no users attached to the kernel via Server manager. During a normal shut down of services the kernels are shutdown with an end controled the problem is if you have a kernel running hot so to say as I just mentioned and the issue to end controled was issued the end immediate command will not work the second time around and that is why operations was then having to issue the end abnormal command.
We had this type of issue it was caused by the way that operations was ending and starting the services.
In short what was happening was the services were started without issueing the CLRIPC command the night before when ending services. The solution I put in place was to do the following. I would had operations move the CLRIPC command to be kicked off before the STRNET command gets issued during startup.
This is all done for us using BMC scheduler. Make sure your shutdown and start up procedures are happening in the proper sequence.
Another reason this could be happening is that when the services are being ended you have a kernel running with outstanding requests eating up CPU and basically doing nothing. A kernel like that will end fine with an end immediate command issued. Just make sure there are no users attached to the kernel via Server manager. During a normal shut down of services the kernels are shutdown with an end controled the problem is if you have a kernel running hot so to say as I just mentioned and the issue to end controled was issued the end immediate command will not work the second time around and that is why operations was then having to issue the end abnormal command.
Similar threads
