DSXchange

If you run the job with a 1-node configuration, or specify "entire partitioning" for the lookup link, do you still get the same problem?

All DataStage wants to know about are the MQ queue manager and teh MQ queue name. If those can be accessed from the DataStage server command line using the mq commands (i.e. "dspmqaut") then you are good to go from DataStage.

I use "cat -v" on UNIX. I know it might sound repetitive - but please check the creation date/timestamp on the output file to ensure you are lokoing at the right one. Also, check the path in the source file in the job to make sure you are reading from the correct place. You shouldn't be getting iden...

Well, that does help a bit with dbx. If you call it up with the core file (and the full path to the executable) then I think all versions will give a stack trace right away and usually that will give some indication of what caused the problem. If you have an active support contract and no intention ...

What a day, thanks Craig. I was thinking "ipcs" but wrote "IPSC" - which is a group that I used to belong to.

Is the path to this executable different from your current installation? If not, gently issue a "kill 20382" - the process is probably tired since it has been running since April 1st.

Actually, before doing that, what is the result "ipsc -a | grep 0xace"?

ascen - after applying the patch, is your error for the 01/01/2500 date still the same, while that for 01/01/2008 results in "authorization failed" or is that the error for both dates?

There is no specified time period to wait. Sometimes you can shutdown DS and immediately bring it back up - sometimes it takes a while to close down processes and sockets. After bringing down DataStage (i.e. "$DSHOME/bin/uv -admin -stop") you should check to ensure that all sockets have been freed a...

Yes, the time need to write messages to the log can slow down job execution very significantly.

Is the exception handler being triggered and the notification activity not working or does the exception handler not get called at all? Have you ticked the "automatically handle activities that fail" box in the job control?

all you need is to have "dbx" installed in order to analyze core files.

It would seem that you aren't calling this as the actual DataStage administrator. I would recommend fixing the problem in the Admin tool and deleting from there; the problem you see might be indicative of other issues. If you running this as your dsadm user, could you please post the whole error mes...

At Version 8 the problem no longer exists. At Version 7 you can use the command "orchadmin rm {path to dataset}" as a UNIX shell command in the before-job routine call for your job. That will remove the file (using the configuration stored in the file itself) and will use the current APT_CONFIG sett...

Does the transform stage do more than a lookup? If it uses stage variables and derivations, then try splitting the lookup functionality into another transform stage. Then see if the lookup part or the other part is triggering the abort. You mentioned that this is a sporadic problem; does that mean t...

The first question in cases like this is "do you have RCP enabled?"