DSXchange

Minimum.

Definitely minimum. Particularly if all tiers (repository, services and engine) are to run on that box.

You can "increment" a single character.

pandujoy wrote:is there a way to just make both (source and lookup) fixed widht

Yes, make sure that they have identical data type and are padded with identical characters.

Is your Shift key broken? You aren't e.e.cummings, so please don't try to be.

Advanced Find allows you to filter on precisely this kind of requirement. Right click on the result of the find to do whatever else you need, such as export, cross-project compare, etc. Advanced Find allows you to search for (wildcarded) text strings in name and/or description, and its filters inclu...

While Windows does not have an .odbc.ini file nor the ODBCINI environment variable, it does have a Registry entry called ODBC.INI (a hangover from Windows 3.1 and earlier when there wasn't a registry). This contains ODBC data source names and definitions. Within any DSN for an Oracle connection expl...

To do N at a time, create N disjoint lists of file names and run these into a multi-instance sequence (as described above, but multi-instance) all controlled from a master sequence.

Nor am I.

IBM (in some countries, at least) does offer a service for implementing HA in DataStage environments but I have never been involved with any such engagement so can not advise on the details of what it involves.

What is reported in the ASB agent's log file?

Who set the setgid bit on OVER.30 ?

What are the permissons on the hidden file .Type30 ?

I too am surprised by that result. Does it obtain when you're using "regular" metadata rather than a schema file? If not, then you've probably uncovered a bug that you can report to the vendor.

True but, for some reason, the OP specified that he/she want's to use the DataStage (processing) node number. In our various ways we have been trying to advise that the two (DataStage and DB2 node numbers) are unrelated. If you know the number of partitions of the table and specify the DB2 node numb...

Is something else (maybe the MKS Toolkit telnet service) using port #23 on your production machine?

Can you telnet using port 2300 (telnet hostname 2300) ?

No, unless you can guarantee that you have accessibility to every kind of database to which DataStage can connect.

<rant>Resist stupid requirements!</rant>

There's more than one way to skin a cat in DataStage. Here's another solution: Hashed file stage to read the directory "keys" (file names) from a pathed "hashed file" (Type 19). Constraint in Transformer stage to extract and compare the date. Command stage to effect the delete.

Separate the reading and the parsing. Read the file as having a single field (data type VarChar), with first line is column headings set so that the first line is ignored if that's appropriate. In a downstream Transformer stage use substring techniques to parse the record into its composite fields.