Page 1 of 1

Error Msg- Cleanup after ABORT raised in stage

Posted: Mon Aug 08, 2005 2:34 am
by ysweec
Hi,

Any idea how to troubleshoot the DS jobs if getting following error:-
DataStage Job 1051 Phantom 2051
Program "DSD.IPCOpen": Line 185, WRITE failure.
Attempting to Cleanup after ABORT raised in stage SMSCERRORLoadFactMetricaSmsError..Link_Collector
DataStage Phantom Aborting with @ABORT.CODE = 3

Regards,
SC

Posted: Mon Aug 08, 2005 2:52 am
by kumar_s
Hi,
i guess this comes at the ouput stage.
may i know what the value set for "Clean On Failure"

regards
kumar

Posted: Mon Aug 08, 2005 3:14 am
by ysweec
Hi Kumar,

I have no idea what is the value set for "Clean On Failure". Can u guide me where to find it?

Regards,
SC

Posted: Mon Aug 08, 2005 3:34 am
by ArndW
Do a job reset and look into the new message(s) in the log file, often they contain additional information.

Posted: Mon Aug 08, 2005 5:57 am
by kumar_s
ysweec wrote:Hi Kumar,

I have no idea what is the value set for "Clean On Failure". Can u guide me where to find it?

Regards,
SC
Hi,
u can find this in the output stage, under properties tab -->Opiton ---> Clean On Failure

regards
kumar

Posted: Mon Aug 08, 2005 7:14 am
by chulett
kumar_s wrote:u can find this in the output stage, under properties tab -->Opiton ---> Clean On Failure
Huh? Maybe in a PX job, but not in Server anywhere that I recall.

The best thing to try is what Arnd suggested - reset the job after it aborts via the Director and see if any messages come into the log labelled as 'From previous run...'.

Posted: Mon Aug 08, 2005 7:56 am
by kumar_s
Hi,
Iam sorry, yes that is in PX jobs :roll:

regards
lumar

Posted: Mon Aug 08, 2005 6:39 pm
by ray.wurlod
Attempting to cleanup after abort is purely informational, and can be noted and ignored.

The actual error has occured in a Link Collector stage, and is a failure to write - presumably to a buffer used by that stage or to a file (Sequential File stage?) connected to its output. This is most likely to be a permissions or space problem.

Reset the job in Director; there may be a "from previous run..." message logged that contains additional diagnostic information.

Also, determine the name of the directory referred to in your UVTEMP configuration parameter, and check that its file system has plenty of free space.

Posted: Tue Aug 09, 2005 9:14 pm
by ysweec
Hi Ray,

Yes. My job is job server.
10x! That job can be reset and re-run successfully.

"determine the name of the directory referred to in your UVTEMP configuration parameter, and check that its file system has plenty of free space. "

Would like to what do u means by "determine the name of the directory referred to in your UVTEMP configuration parameter"? The name of the directory in where?

Regards,
SC

Posted: Wed Aug 10, 2005 1:13 am
by ray.wurlod
In the DataStage home directory is a file called uvconfig containing a heap of DataStage configuration parameters. UVTEMP is one of these. Its value is a directory used for various kinds of temporary file by the DataStage run machine.
From the operating system you can execute a command such as

Code: Select all

`cat /.dshome`/bin/analyze.shm -t | grep UVTEMP

Posted: Wed Aug 10, 2005 2:10 am
by ysweec
Hi Ray,

10x so much.
So, u means just check the UVTEMP is refered to which directory and check the available space for that directory, right?
Anything to tune on the DS jobs?

SC