DSXchange

We're way past the DSJ.ME issue and into the 'interlude' now. Please try to keep up.

Ignore him.

A combination of Craig and Kumar's suggestion. Change your job design to <snip> Can't work that way, all the metadata would have to be the same for the keys in each table. Unless you mean four sql statements in the 'before' tab and something bogus in the main sql area. Yeesh... Use four links. Easy...

There's no files, dummy or otherwise, involved with the 'source from a Transformer' method. Just literally four Transformers stages all linked to one OCI stage. There are no input links on the Transformers, only one output link apiece to the same OCI stage. Create a bogus stage variable in each tran...

One Oracle stage feed by four Transformers, each providing the keys for the delete statements. All kinds of discussions here on how to 'source' from a Transformer, you pretty much just need a stage variable and a constraint so it does not feed in rows forever.

dir /b | find ".zip" should work just fine from DOS as long as you are sure you are in the right directory. Not sure what else the issue might be.

Or just do dir /b *.zip instead.

Exactly. The 9i Client software should be installed and configured on the DataStage server and then you can use that to connect to any 10g instance.

Reset the job after it aborts and post any 'From previous run...' log message.

You mean a shared or local container? I don't think that matters. Are you specifying an active stage for the function?

Ugh. Or just use a tool designed to move Mainframe/UNIX files at high speed and do the EBCDIC to ASCII conversion on the fly during the transfer.

Yes - OCI9 and OCI Load both, with a 9i client. They don't 'officially' support 10g until 7.5.x from what I recall.

Syntax dir /a-d|find ".zip" It is giving me an error saying Result = find: unable to access ".zip": The system cannot find the file specified. The same syntax works fine from the Command Prompt... You are probably not in the directory you think you are in. Are you using fully qualified paths or at ...

It should mean you left out the connection information. Are we talking Oracle here? Then your script needs to spawn a sqlplus session, supply the logon credentials it needs and then run your sql script. Would be pretty much identical for other databases except for the 'sqlplus' bit. That's what any ...

As noted, please let us know under what circumstances you'd like to be able to skip it. Otherwise, a generic answer would be to branch around it using either Triggers or a Nested stage and then a Sequencer set to 'Any' to bring the stream back together before Job3.

Good to know! Thanks for posting that...