DSXchange

Not sure what 'best' would mean here. Discuss both with your IBM rep and see which sounds best given your goals for this. Going to 8 is no small undertaking, btw.

OK, for the 'collect the stats after each job run' approach, ETLStats gives you a huge leg-up there.

Doesn't really change my previous answer - you can go to the last 7.x version (7.5.3) or the latest 8.x version (8.1). Both are available in Enterprise Editions, if that's what you want.

Are you looking to do this identification after the fact or as the jobs are running? Regardless, trying to determine how long is 'too long' is not always a simple task.

OK, how about some details of your job design? Anything running in Sequential rather than Parallel mode?

How much of an upgrade are you contemplating? In my mind, you have two choices: 1) 7.5.3 to get to the last 7.x Server release. 2) 8.1 which is the latest 8.x release but is a completely different architecture. I'm sure IBM or whomever is your vendor for that would love to discuss the gory details w...

What kind of partitioning are you doing in the job?

I think you were plenty clear. Anything can set environment variables, the only thing setting them up in the 'admin tool' does is to make them specific to the DataStage runtime environment but not otherwise generally available. Keep in mind the fact that you'll need to get the value of any environme...

You'd need to use that in a BASIC Transformer as 'ereplace' is a Server function. However, there is a PX version posted here somewhere.

You really need to start your own post rather than hijack someone else's "similar" topic.

Index() will find it. Simple trick to match the whole word is to concatenate your delimiter to each end of both the string to search and the string to search for: Index(" ":YourField:" ", " BANK ", 1) That will match BANK (even at the ends) but not, mmm... BANKRUPTCY, ...

Is it a secret?

Two things you should really do now. Post the resolution so others coming here with the same problem will know what fixed it for you and mark the post as Resolved using the big button at the top of the page.

I don't really see a problem with job control that knows what the definition of those variables are and can set them properly as environment variables before launching a job that leverages them. You would use the default value of $ENV for them in the job. Or if you are using 'static' files, it would...

By dividing the data up across the 'nodes' by partitioning.

That's not the only error in the log, is it?