DSXchange

We are getting the following message and some of the jobs are being aborted by giving one of the below warnings

Abnormal termination of stage LdPgmRlsGmew.6575.xLoad detected
Job Aborted after 1 errors logged.
Attempting to Cleanup after ABORT raised in stage LdPgmRlsGmew.6575
Job control process (pid 791) has failed
Job Aborted after 1 errors logged.
Attempting to Cleanup after ABORT raised in stage SeqPgmRlsldGmem.6575.JobControl

OR

SeqPgmRlsldGmem.6568.JobControl (@CmprVppsLvl3): Controller problem: Error calling DSRunJob(PgmRlsLdGmewCmprVppsLvl3.6568), code=-14
[Timed out while waiting for an event]



We have server jobs and parallel jobs being run on the same datastage server and we have heavy usage of mulitple instances in server jobs.

When i re-run the job, job finishes successfully. I am not getting any informative message in the log to debug the cause of this occurance.

Is it running out of shared memory ?
If so how can we avoid it

i see there is a Ecase for increasing the timeout , but does it resolve the issue we are having???

The (-14) [Timed out while waiting for an event] issue is quite common and has been discussed for years here. It's something to be "managed", but not resolved, from what I have seen.

Just yesterday Arnd talked about a custom patch written for his specific customer to extend the timeout value, but I've never had that success for my customers.

Just an additional Info

This is the log from previous run when i reset the jobs for the Abnormal termination of stage.

From previous run
DataStage Job 1607 Phantom 791
Job Aborted after 1 errors logged.
Program "DSD.WriteLog": Line 201, Abort.
Attempting to Cleanup after ABORT raised in stage LdPgmRlsGmew.6575

DataStage Phantom Aborting with @ABORT.CODE = 1

kcbland wrote:The (-14) [Timed out while waiting for an event] issue is quite common and has been discussed for years here. It's something to be "managed", but not resolved, from what I have seen.

Just yesterday Arnd talked about a custom patch written for his specific customer to extend the timeout value, but I've never had that success for my customers.

I have see the patch, but i would like to know if the patch really worked for any of the customers.

Any views on the abnormal termination of the stage???

What kind of stage is it, and to what kinds of stages does it connect?

Without that information your question is in the "my car doesn't work, what's wrong with it?" category.

ray.wurlod wrote:What kind of stage is it, and to what kinds of stages does it connect?

Without that information your question is in the "my car doesn't work, what's wrong with it?" category.

The abort is happening on Transformer stage. source is a hash file and target is a flat file. Between the transformer and target there is a sort stage.

The "patch" is an Ascential one and the tip did not come from me (I was the one looking for help) but from Klaus Schaefer. Click here for the link

Hello there,

Could you let me know from where the patch can be downloaded?

Thank you very much for your help

ArndW wrote:The "patch" is an Ascential one and the tip did not come from me (I was the one looking for help) but from Klaus Schaefer. Click here for the link

www.ibm.com and then go to support. You will need your IBM login id and password and can then search on the eCase/patch number and download from there.