Abnormal Termination Issue
Abnormal Termination Issue
This is a topic for an orphaned message.
Hello
I too have unexplained job aborts lately off and on. Once I either reset or recompile the job it will run fine. There is never enough about the error either. I have even checked the &PH& log and nothing more there helps. I have been working with Datastage support and they believe the
project temp space could be an issue. I am working on cleaning this out as
much as possible and testing this suggestion. Does anyone else have any suggestions?
Stashima
I too have unexplained job aborts lately off and on. Once I either reset or recompile the job it will run fine. There is never enough about the error either. I have even checked the &PH& log and nothing more there helps. I have been working with Datastage support and they believe the
project temp space could be an issue. I am working on cleaning this out as
much as possible and testing this suggestion. Does anyone else have any suggestions?
Stashima
You need to determine if you have run out of virtual memory when youve gotten the abnormal terminations. The easiest way to see this is to watch top for the swap free value going down to near zero. If your job designs are using large hash files, with read or write caches, you use up memory. If the free space drops to zero, you MAY get messages in a job about unable to allocate heap. Meanwhile, any other jobs executing could get abnormal terminations.
By the way, if youre on DataStage 3.6.3 on Solaris 2.6 using Suns Photon drives, you need to get your hands on some patches Sun had to write in an emergency situation last year, and also get some patches from Tech Support. I had to escalate this problem to tech support for a client who had that configuration and it turned out Sun has a bad bug in 2.6 that causes abnormal termination messages in jobs whenever cpu and disk utilization reach peak levels. It was some timings in the OS and hardware.
Good luck!
-Ken
SHailey@CSCUS.JNJ.com on 17-Sep-2001 15:55
Please respond to datastage-users@oliver.com
To: datastage-users
cc:
Subject: RE: Abnormal Termination Issue
Hello
I too have unexplained job aborts lately off and on. Once I either reset or recompile the job it will run fine. There is never enough about the error either. I have even checked the &PH& log and nothing more there helps. I have been working with Datastage support and they believe the
project temp space could be an issue. I am working on cleaning this out
as
much as possible and testing this suggestion. Does anyone else have any suggestions?
Stashima
By the way, if youre on DataStage 3.6.3 on Solaris 2.6 using Suns Photon drives, you need to get your hands on some patches Sun had to write in an emergency situation last year, and also get some patches from Tech Support. I had to escalate this problem to tech support for a client who had that configuration and it turned out Sun has a bad bug in 2.6 that causes abnormal termination messages in jobs whenever cpu and disk utilization reach peak levels. It was some timings in the OS and hardware.
Good luck!
-Ken
SHailey@CSCUS.JNJ.com on 17-Sep-2001 15:55
Please respond to datastage-users@oliver.com
To: datastage-users
cc:
Subject: RE: Abnormal Termination Issue
Hello
I too have unexplained job aborts lately off and on. Once I either reset or recompile the job it will run fine. There is never enough about the error either. I have even checked the &PH& log and nothing more there helps. I have been working with Datastage support and they believe the
project temp space could be an issue. I am working on cleaning this out
as
much as possible and testing this suggestion. Does anyone else have any suggestions?
Stashima
Try checking the &PH& log :
1.) before the job abort( if possible)
2.) just after the job aborted (before resetting the job)
3.) after resetting the job
Regards,
William Tan
--------------------
----- Original Message -----
From: "Hailey, Stashima [SLC]"
Date: Tuesday, September 18, 2001 3:55 am
Subject: RE: Abnormal Termination Issue
> Hello
> I too have unexplained job aborts lately off and on. Once I
> eitherreset or recompile the job it will run fine. There is never
> enough about
> the error either. I have even checked the &PH& log and nothing
> more there
> helps. I have been working with Datastage support and they believe the
> project temp space could be an issue. I am working on cleaning
> this out as
> much as possible and testing this suggestion. Does anyone else
> have any
> suggestions?
> Stashima
>
>
>
>
1.) before the job abort( if possible)
2.) just after the job aborted (before resetting the job)
3.) after resetting the job
Regards,
William Tan
--------------------
----- Original Message -----
From: "Hailey, Stashima [SLC]"
Date: Tuesday, September 18, 2001 3:55 am
Subject: RE: Abnormal Termination Issue
> Hello
> I too have unexplained job aborts lately off and on. Once I
> eitherreset or recompile the job it will run fine. There is never
> enough about
> the error either. I have even checked the &PH& log and nothing
> more there
> helps. I have been working with Datastage support and they believe the
> project temp space could be an issue. I am working on cleaning
> this out as
> much as possible and testing this suggestion. Does anyone else
> have any
> suggestions?
> Stashima
>
>
>
>