DSXchange

Hashed files are insert only. Destructive overwrite does all the rest.

Stage variable.

Is your target database going to have any clue what to do with the DataStage routine? Nope. Ok, let's qualify that - no, you can't use it directly in the query. You could, however and for example, execute the routine outside the job, pass the value in as a job parameter and then use the job paramete...

I don't think anything implicitly converts in parallel jobs.

I found his omnisciency interesting as well. Assuming it was a 'been there, done that' kind of moment for Chuck.

Not sure how it's 'much simpler' as it is practically identical to the previous method. The 'dummy' file would have the same contents as what would be echoed, so basically the only difference would be: type dummyfile >> #FileName# -versus- echo 0000000000000000000000000000 >> #FileName# Use 'cat' on...

It always helps to mention the database involved as that can change the answer given. That being said, what is your doubt? Based on what you have written it sounds very straight-forward. Select the keys and data from Table A and send them straight to Table B with the action set to 'Update existing r...

Good to know they are 'way fast'.

I was thinking about cutting the string up around the pairs and putting the chunks together, but if iterating through character by character is speedy - you're right, that seems the simplest approach.

I have a field I need to strip pairs of characters from, basically any place I find a caret ^ in the string I need to remove it and the immediately following character. Problem is it's not a small list like only '^R' and '^C', there are 90 different values so something simple like EReplacing them wi...

Interesting. I'm still on 7.0.1 so perhaps that's something 'they' have fixed. Sure would simplify things.

naveendronavalli wrote:I was thinking of putting the dummy record in an other SEQ FILE and then appending to the written SEQ FILE. However, this sounds to me a little inefficient way of doing it.

Nothing inefficient about it and sounds like a perfectly fine solution. I'd say go for it.

jobstatus was intentionally removed. However, I have updated the code which includes the jobstatus with the same problem. A couple of things to keep in mind... The -wait and -jobstatus commands are mutually exclusive as both tell it to wait before moving to the next step. You should use one of the ...

Welcome. :D Search the forum. The concept of DataStage running clustered or at least in a clustered environment has been discussed several times in the past. It really depends of what you mean by 'run in clustering'. While it can certainly run on a cluster, and can be made to fail-over from one node...

But first you've got to read it into the job and those invalid inner quotes will make that... difficult.

As noted, lots of good information here - but it sounds like you should stick with your original approach, Arnd. I've worked with someone in the past to do something very similar to this and it worked out rather... nicely.