Page 1 of 1

Abnormal termination of DataStage -- Fault type is 6

Posted: Mon May 02, 2011 11:14 am
by swades
Hi ...

I searched the forum but could not find any solution...

We have a master seq. job which invokes 35 different child DS job in DS v8.1 FixPack1 environment (OS - Red Hat Linux v5.5).

master seq. and all child job (both) are multi-instance. master seq has to run 25 times with different parameter to complete full data load.

One of the datastage job failed twice yesterday with below error (same job run fine without any modification in second attempt)

Job control process (pid 406) has failed


job code::

ODBC connector -- > transformer --> Oracle connector stage

ODBC connector stage used to fetch data from DB2 running on AS400

and this is what I got in director after reset the job

From previous run
DataStage Job 228 Phantom 406
The variable "APT_PERFORMANCE_DATA" is not in the environment.
terminate called after throwing an instance of 'LoggingAgentConnectionException*'
Abnormal termination of DataStage.
Fault type is 6. Layer type is BASIC run machine.
Fault occurred in BASIC program DSD.WriteLog at address a3c.


What is meaning of "Fault type is 6"?

Thanks in advance.

Posted: Mon May 02, 2011 2:08 pm
by ray.wurlod
Fault types are documented in an appendix of the UniVerse Administration manual.

Posted: Mon May 02, 2011 3:06 pm
by swades
ray.wurlod wrote:Fault types are documented in an appendix of the UniVerse Administration manual. ...
Thanks Ray!

Any idea :idea: about this error?

Posted: Mon May 02, 2011 4:24 pm
by ray.wurlod
Not without consulting that manual, which I'm not in a position to do right now.

Posted: Mon May 23, 2011 9:58 am
by swades
Hi Ray,

This is what I got response back from support,

The abnormal termination of stage message indicates the writing of the log failed causing the termination.

Please refer to the following technote to change the logging, as described:

https://www-304.ibm.com/support/docview ... wg21370048

Change to
RTLogging=1
ORLogging=0


Is this make sense to you? not to me :cry:

Posted: Mon May 23, 2011 2:37 pm
by ray.wurlod
Yes, it makes sense to me.

You had logging set to perform logging into the metadata repository. Something went wrong with that - it seems to have been being accessed via (an internal) Connector.

The change that Support recommended switches job logging to the engine tier, using the old style of logging into hashed files.