DSXchange

The "list of errors and recommended actions" manual is promised for the next ("Hawk") release. It does not currently exist as a single document. Some error codes can be decoded via the SYS.MESSAGE table in DataStage. This has six digit (character) keys: SELECT * FROM SYS.MESSAGE WHERE @ID = '081011'...

Not so much "overhead of BASIC" as "overhead of translation into and out of BASIC (typeless) environment from the C++ (strongly typed)envionment". It is likely that this overhead will cause perceived slowness in server jobs in Hawk, because the same transitions need to occur (everything operates in ...

Help for the job sequence properties is clear on how the compilation options work. It depends on whether an Exception Handler exists and on whether explicit handling (what you used to have) exists.

... and being granted permission to schedule on UNIX (cron and at)

It's because you have not done the binding of DataStage roles to Windows groups as described in the Administrator guide.

17 lightweight things might load the system less than two heavy-duty things.

Server jobs do not enforce data types or nullability. Therefore, even though the Key column is marked Not Null, you are not prevented from sending NULL. It is up to you to prevent this from happening. Or switch to parallel jobs, where data types and nullability are enforced.

Not concurrency, just total load. Monitor your operating system to determine where the bottlenecks are; CPU, memory, I/O throughput, network bandwidth are the primary candidates.

cron running dsjob (via rsh)?

Hashed files don't have data types. Everything is stored as a delimited string. What you store is what you get.

Did you "find" that solution because Craig had posted it a mere 15 minutes before your post?

Folks using 7.5x2 clients to access 7.5.1A servers, for which combination the Palette is horked.

or

or just

You don't have to add it to your Palette unless you want it there. You can use it directly from the Stage Types branch of the Repository.

If it can't open the hashed file, then the improper data type may have resulted from an attempt to use a file variable that relates to a file that is not - or is no longer - open. This can occur, for example, if the hashed file was opened successfully when the job began but was force closed or delet...