Stopped job runs in background

[JKenklies]

Hi folks,

I have made an error last week. I've made a lookup on a DRS stage and tried to use a user defined SQL. That resulted in getting the whole lookup table for each input row, so I got the Error

Code: Select all

Reference lookup generated <n> rows. Only the last row was used.

In this case, n was quite huge, the whole table (about 4,5 Million rows) has been fetched for each input row (again some millions).

Well, I noticed that error and reversed my changes. I stopped the job, recompiled it and ran it without getting that error again. Today i reviewed the job in cause of other changes and noticed that in the job log, I'm still getting that Error with the huge number of generated rows. Our DB2 Admin sais, the impact is not critical, but I think thats only because the buffer pool sings the same song again and again...

I think this is a kind of phantom job. I'm not able to stop it, because the stop-button is disabled.

Q:
How do i stop this SQL-Fireing process? Will I have to restart the ETL-Machine? Is there a smarter way to get rid of this?

Thanks in advance.

[kcbland]

Make sure there isn't an active thread still running, like a zombie. Of course, rebooting the server is like a Microsoft solution but it just may be the easiest.

[JKenklies]

I solved the problem. I've used sysinternals process explorer to be sure which process I had to kill. It shows the command line that has been used to start the process, so I could see the job's name.