DSXchange

You can use the hashed file stage to feed a reference input link to the Transformer stage.

You cannot use a DRS to access a DataStage hashed file to feed a reference - or, indeed, a stream input link.

You can use a DRS to access a database table to feed a reference input link.

But if you've exported the job executable, you don't need to recompile.

How? You have to be in Designer to create/modify the job designs and you can compile from any client (multiple compile tool). How does being able to do so from a command line add anything?

On Windows you will need to get MKS toolkit or something similar to use the wc command. DataStage counts the rows that it processes. That's one source. For the other, create a small DataStage job that executes SELECT COUNT(*) FROM whatever and returns just that number into a file. You can then easil...

0xB3 is Char(179) which is not a printable character in ASCII (which is what DataStage uses if NLS is not enabled). It will handle the character OK, but will have difficulty displaying it. It will almost certainly not be able to display it as superscript given that you're on a UNIX server; UNIX-base...

You might also check the indexing in the project, since the helper subroutine was DSR_SELECT.

Probably the easiest, and closest in intent to Oracle's ROWNUM, is CALC EVAL "@NI" which counts the number of items.

Or CAST your some_date column as DATE so that its type corresponds with the type of the constant CURRENT_DATE.

The less things stay the same, the more they change.

Welcome aboard! :D

And thanks for the useful advice. Obviously if you send (or request) a large array of information to (or from) the database, it will need more memory to store the array than it would for a smaller array. And you've hit a configured limit when this error occurs.

DRS does not talk to DataStage hashed files. The only way that the DRS can talk to UniVerse hashed files is if you have a UniVerse ODBC driver installed on your DataStage server machine.

In particular you can not query the Repository to get schedule information; it's simply not stored there.

Erk!

As a quick fix, since it only requires three columns, what happens if you actually create and populate a MetaStage_Loc_Info table?

Are you running SP2? If so, have you installed the SP2 patch from Ascential?

If you "delete the last N entries" then you won't be able to see just "the last couple of entries in the log". Of course it is possible to purge up to the previous run, and it is recommended that you do so, either manually or automatically. More exact purging can be done using DML, but this is not r...

You could even create little
DataStage jobs to insert/update the table, and have them run from before/after subroutines or, indeed, from a master job sequence. This may be easier for you to maintain.