DSXchange

Hi there,

I am running DataStage 6 on Solaris 8. I had 9 good runs and suddenly on Saturday my jobcontrol started to freeze up. It showed that the job is still running but nothing happens.
Where one of my jobs normally extracts data at +/- 3500 to 4500 rows/sec on Saturday it extracted at 33 rows/sec.
DataStage was up and running but when i tried to go into one of my projects via Designer or Director it just never returned after inserting the username and password.

After restarting the DataStage server this morning, I tested my job and everything ran just fine again.

Any suggestions would be appreciated.

Thanx.

Faeez

Hi Fkana,
i suppose that DataStage log file are corrupt.

try to execute this unix command

cd <ascential project directory>
for i in RT_STATUS*
do
echo $i
<ascential direstory>DSEngine/bin/clear.file $i
done

this command clean the log file

regards

Mario

Well... keep in mind the fact that this will totally clear all of your job logs. And by 'clear' I mean not like you would from the Director, but totally blow away everything in them, including any Auto Purge settings you may have had.

There really isn't enough information to make a judgement regarding what may or may not have been happening. Could have been related to DataStage or the server itself or something else running on the server that sucked up all of the resources...

Hi Craig,

Will a project cleanup do the trick?

Ciao.

Faeez

Hard to say... couldn't hurt!

A project cleanup will delete unreferenced files, etc. but will not clear log files.

Faeez, if it happens again, why not check your running processes and CPU/memory resources before restarting either the DataStage server process or the machine itself? I had a case once - it wasn't a DataStage project - where the same sort of thing happened. It was thought that the MOLAP server we implemented was behaving weirdly, but it turned out that an anti-virus process was vehemently trying to access and scan a MOLAP cube file that was opened and locked by the MOLAP server process. The two processes ended up 'fighting' over the file and the anti-virus process was using 99% of the resources, making the MOLAP server as good as dead - it was running, but for all practical purposes did nothing. Of course it seemed to solve the problem when killing and re-spawning the MOLAP server, because the anti-virus of course got its way, but it was the anti-virus's fault!

I'm not saying that this is what's wrong in this case, but just check... a Saturday also sounds suspect - who knows what's scheduled to run at such a quiet time?

Thanx guys.

Faeez :D