DSXchange

Vineet Pal Singh wrote:Could you please provide some pointers on how to do it?

I actually left that as an exercise for us1aslam1us to complete, but it seems Eric beat him to it.

It can all be done by using a stage variable to store the previous value and populate this vale in the current Column. Just wanted to point out that simply saying you can do it doesn't help someone who doesn't know how to do it. Best to flesh out an answer like this with an actual example of how on...

Yes, always better because NOTFOUND doesn't really work for non-hashed file based lookups.

Hard to tell, though - is your lookup to a hashed file or directly to a database table?

Can't say. From what I understand it is basically equivalent to "en-US" and mine is set the exact same way (LANG=C) on a (non-DataStage) Sun system. For whatever that is worth.

Resource contention. More != faster/better. Sometimes also known as the Law of Diminishing Returns.

You should have led with that, this has nothing to do with "access" in the sense we were all thinking you meant. This just means that your $DSHOME/bin directory (where the command lives) is not in your PATH statement. So, either update your PATH in the dsenv file to include that directory ...

deepu_n55 wrote:i am using ODBC table definition here in the OE stages.

Don't. They're not necessarily compatible. Import the metadata using one of the Oracle mechanisms, use that.

Well crap, could have sworn I took care of that already. Moderator powers activate! :D

(feel free to edit the title if it doesn't suit you)

Lenny - Reply to topic not Reply with quote. Please. Learn it. Love it.

And yet you haven't marked it as resolved, so... is it?

I don't understand the point of using a sequence generating object like the Key Management routines if you need to restart it at 1 each time which defeats the whole purpose. Just manage / increment it yourself.

Please clarify for us exactly what your issue is, as it is a little too nebulous right now. Are you seeing an error? What happens when you try to execute dsjob?

As stated in the Tech Notes, do NOT do that if ANY jobs are running. And don't confuse the "size" of a directory with what you are seeing with the log's hashed files, they really aren't equivalent.

Ack... or that, I answered too quickly. Depends on if you need to run a specific job or jobs from a routine or if you need to run one job for every row passing through a job, which honestly doesn't requite a routine. Ray's answer is more appropriate for the question as asked...

Search here or in your documentation for the UtilityRunJob function.