DSXchange

Please delete the other two copies of this post (one is in the parallel forum). 1. Yes, if the need is there. Depending on the total volume of data to be loaded into the hashed file, you may need to create/resize it to use 64-bit internal addressing. 2. While loading, pre-tune the hashed file to the...

Accumulate the entire message in a Transformer stage, follow with an Aggregator stage that passes only the Last in each group (message ID).

Message headers, presumably. What does the application that is writing to the queue do to identify whole file versus part file message?

Problem 1 - move the OPENSEQ statement outside of the loop. Problem 2 - The isolated 1 is being treated as a format specifier. Lose it. If Len(stage_variable) Then stage_variable : @VM : inputlink.column Else inputlink.column If these aren't the answers you need, post your actual code, between Code ...

And that (port #21) is what the FTP stage is trying to use.

Write a script to use secure FTP and invoke that from DataStage.

In Director Schedule view or through inspecting the crontab entries for the DataStage user.

You should get one process with your netstat query. This is the dsrpcd process - it should be there, with a state of LISTENING. You can shut down gracefully if this is the only process on port 31538.

Does it simply not get focus? Minimize all other windows (including Designer) to see whether this is the case.

Not to Rejects, but perhaps to an Otherwise/Log output link. Set up your constraint on the main output so as to prohibit nulls.

Welome aboard. Set the job sequence job properties option that generates a warning when an activity fails to finish with status of Success. This will propagate warnings upwards.

So what precisely is your problem? Sounds like you need to initialize the SDKSequences hashed file within the job sequence (batch). Search DSXchange for details of how to do so.

What have you tried?

The answer that most readily springs to mind is to use a UniVerse stage (which means that you must have a VOC pointer to your hashed file) and in it derive SGKey as MAX(SGKey).

That is so wrong. Status 1 means OK, status 2 means that there was at least one warning/alert message logged. But we're straying off topic.

Pedantically, perhaps, I regard a join specified in a database stage as not being a join done in DataStage - it's done in the database server.