DSXchange

The process table is a low level component in UNIX (think of the pid as a row number in this table). There is one entry associated with each process; it's used to keep track of timeslice, execution priority, and so on. The fact that you can execute when not writing to a table seems to exonerate the ...

The Transfomer stage with stage variables approach will work irrespective of source and target. The only requirement is that source is sorted.

No. If you do that you end up with orphan tables (though these can, in theory, be cleaned up from DS.TOOLS menu). Worse, you get lots of orphan records in DS_JOBOBJECTS, and the DS.TOOLS utility does not (as of version 7.1) clean these up. Better is to figure out why RT_CONFIG740 is giving problems....

Yes, good catch, I don't often get to worry about the physical.

The "other" name, usually "mounted on" in the df report, is the one I normally use.

Without knowing the exact details of what you are trying to do, the first thing that jumped into my mind was Merge stage.

Welcome aboard! :D You basically have two choices. Selecting NULL from a database does not present any problems; it's entirely legal to do. However, there is a mapping function in all database SQL for converting from NULL to something else. I don't know what the function called in DB2 - it's called ...

You could pass the data through a QualityStage UNDUP operation. A hashed file is a good way to remove duplicates based on key value, but is not appropriate otherwise. If you are contemplating the need to remove duplicate rows, sort the input (use sort as a filter in the Sequential File stage) and us...

When you tried to import the table definition from the ODBC DSN, did you have the "include system tables" check box checked? Each worksheet in a workbook is reported for some reason as a system table. It is essential for Excel ODBC to work that the worksheet is in a regular tabular format with colum...

Look in your job log and advise what the output from the SETFILE command was.
This may indicate the source of your problem. Hashed file names and VOC entry names are case sensitive and must be spelled correctly.

You won't be able to attach the job log, because the job log is a database table. Therefore you're up for writing your own routine to dump the contents of the log into a text file. In which case, you may as well create a routine to send the email, which can invoke the "dump" routine, and you can be ...

What would you expect to get if the job status is other than finished?
Think about what the other possibilities are: still running, aborted, finished with warnings, crashed, stopped (by operator).

DataStage saves your job designs in the project. But you are permitted to create projects anywhere in the file system apart from the root directory. So, if /dev/vg01/ardent is getting crowded, why not create another project on another file system, and do new work there? Or even export everything fro...

I would be very surprised if what you claim is true. Substring is one of the most fundamental functions in ETL data processing.

What evidence do you have for your assertion?

Unfortunately I can not check, as I am currently working on a 7.1 site.

Do you have INFORMIXDIR and INFORMIXSERVER environment variables set correctly in the dsenv file? Is there an entry for the database in $INFORMIXDIR/etc/sqlhosts and does it specify onsoctcp as the protocol and the correct port number? Is that line in sqlhosts correctly laid out? (The error message ...

Reducing the transaction size does help, though the only safe size is 1. Otherwise all you've done is to reduce (not eliminate) the risk that you're trying to update the same page as the one from which you're selecting. This happens in databases that use page level locking rather than row level lock...