DSXchange

Some combination of Date/String conversion functions with appropriate date picture arguments.
Stage variables are OK too; they become local variables in the generated code.

The DSSendMail() function - which is invoked by a Notification Activity - expects an SMTP mail client.

You can use SMTP to send to an Exchange address that automatically forwards it, or you can write your own interface using the Microsoft API.

Since UtilityAbortToLog is written as a transform function, it can not be used in parallel jobs unless from a BASIC Transformer stage. I am even more strict than Kim; I construct jobs that never abort; instead they log warning messages and, possibly, propagate their status to their callers using the...

If "mate" not stated, "mate" is deemed to be implicit, mate.

Please post the full output from the SQL query. Please also post the full output from LIST DICT HOLDER_UV so that you and we can verify your assertion about the data type. If you wish, please also post the output of SELECT COUNT(*) FROM HOLDER_UV WHERE PRODUCT_ACCOUNT_NUMBER = '184306' EXPLA...

Your original post specified DataStage 5.x as the version; I paid more attention to that than to the subject.

As far as I am aware, dscc.exe does not exist in version 5.x. That was the basis of my earlier assertion.

No wucking furries.

There are some SDK Routines for working with dates based on 1900-01-01. Check them out.

You need to establish where (in which folder) the named pipes used to communicate with SAP are created, so that you can check that they are removed once the job completes. Use an Explorer search, perhaps for files modified in the past 24 hours. Why only for SAP BW load? Because nothing else you're d...

The client components, at least, need to be installed on the same machine on which the DataStage server resides.

It is not necessary that the actual JDE, Siebel, etc., applications also be co-resident.

That's because authentication is recorded as fully qualified names (for example domain\name or machine\name) in some locations. There is a hashed file called UV.LOGINS in the UV account that you can use to map domain names, machine names and unqualified names, but there is no DataStage documentation...

Provided your metadata management has been disciplined you should be able to select the hashed file's table definition in Manager and perform a Usage Analysis. This will show you everywhere that the table definition has been used in the project.

Get to a TCL prompt on the server or use the Administrator client's Command window and execute the query select count(*) from HOLDER_UV where PRODUCT_ACCOUNT_NUMBER='184306'; It is syntactically valid. Post the result that you receive - all of it. What you posted is the column heading genera...

If the rejected rows are captured via a reject link, there is an event logged in the job log containing exactly the information you want. You can use Director's "Print to File" capability to write this event to a file, or you can create an after-stage/job routine to read that log entry and write its...

It may use cron or at on UNIX, depending on whether it's a repeating schedule or a "one off" (and on whether the particular UNIX has at or not).

You've used two different casings of the function name (DSSENDMAIL and DSSENDMail) in this thread, neither of which is correct. The name of the function, correctly cased, is DSSendMail.

In the Routines branch of the Repository is an example of how it should be invoked. Take a look at that.