DSXchange

RAI ROUSES wrote:The Job is executed by the Job sequencer, is indicated in log that the Job was initiated, but after same minutes show an error message indicating that the Job was not initiated in 60 seconds.

This is the assertation that my comments are based on. (emphasis mine)

Ugh... I certainly hope not!

7.5 is not what I would consider the 'latest version'... it does seem to be a non-issue in 7.5.2 however. The patch is required up to and including version 7.5.1A.

Worked just fine for me, once I got past the somewhat confusing installation issue mentioned. [shrug]

So it's not really resolved, then.

Perhaps you can open a case with Support, see what they have to say on the issue and then let us know what you find out?

That should address the extract speed. Your next issue will then be how fast can you shove it into the other table. You can follow DSGuru2B's advice at that point.

The result you get in TOAD is actually the throughput. Actually, the result you get in TOAD is fairly meaningless, pretty much only telling you how much time will elapse before the query starts returning rows. However, if it 'runs' in TOAD it will run in DataStage assuming the two queries are actua...

Welcome aboard, Rai. The full text of your error is something like this, I assume: Controller problem: Error calling DSRunJob(JOB NAME), code=-14 [Timed out while waiting for an event] If you search the site for either the error number or the phrase in square brackets, you'll find many discu...

My bat file is used to del those txt files. But that bat file doesn't work properly. How can anyone help you will so little information posted? What would your mechanic say if you called and told them that your car "doesn't work properly", what kind of advice could he give you over the phone? Post ...

Converting a fixed width file to a delimited file would be trivial in a DataStage job.

Yes, you can. It's a wee bit of a PITA but it certainly can be done. There are many postings with the gory details - VOC record, CREATE.INDEX, BUILD.INDEX, creating a hashed file versus a UV table, composite index versus multiple singleton indexes, yada yada yada. I avoid the need to do that unless ...

I agree - silent mode is not the way to go. The sqlplus session isn't all that good about handing back a meaningful exit status. We always redirect sqlplus output to a file in /tmp with the current pid in the name to ensure uniqueness. It is then grep'd post session for things like 'ORA-' or 'SP2-' ...

Ah yes... the dreaded 'negative cool' error.

Sounds like 'patch' might actually mean 'upgrade' in this case.

Subject say export, body say import - which is it?

Can make a big difference, especially from the Server side.