DSXchange

Sounds like someone did a DS.REINDEX ALL with active connections to the repository. Depending on what the answer to Ray's questions are, running it again in isolation may be all that you need.

No worries.

It's a known "issue", no clue why it does what it does, but it does.

As would the Index() and Right() functions.

That's a "problem" your DBA needs to solve by adding more rollback space. That or perhaps you need to commit more often in your job.

That's a new one on me.

The next run will always be "from scratch" unless you specifically code it to be otherwise.

So either you or whomever is building those XML files will need to break them up into smaller deliverables. The stage can process huge numbers of small files but not huge files.

So I am completely lost on what it is you are after in this thread. What "process" are you asking about? Rather than transferring files from one, do you now want to know how to install and run DataStage on a UNIX server? :? You would need to purchase a license for that particular flavor of...

Which is what you have, so you're good there. The issue is the fact that there's no space in front of the AM/PM, not sure why it wants to run the two together but it does. As far as I know, you need to split the result and insert the space yourself, which is easy since the output is a fixed size.

32bit operating system? Your lookup data was too large to be cached in memory.

Set up 50 new columns and use Field() in the derivation of each with the appropriate field number. Some will be null, filter them out post-pivot.

You'll need to split those values out into individual columns and then pivot.

Too may variables with your current approach. Why not do the delete / truncate as a separate operation before this job runs?

Check out the use of the IsValid() function, it can be used for various data types including dates.

Patch.