DSXchange

Expressions in DataStage BASIC may use any of three quote characters - single-quote, double-quote or backslash. Each is correct. The same is not true in DataStage SQL, where specific quote types are required in particular circumstances. In UNIX, too, double quotes are known as "soft" quote...

BCI = BASIC (language) Call Interface, the mechanism server jobs use to mimic the functions in the ODBC API. DSD.BCIOpenR is a routine invoked when an ODBC connection is opened for reading (for example from an ODBC or DRS stage). SQLNumResultCols() is an ODBC function that retrieves the number of co...

Be more specific about what you meant by "didn't work". What came through? Was RCP enabled throughout the whole job?

In my opinion VarChar ought not to be padded. What does your official support provider say??

Easier approach. Use a standard fork-join design, with an Aggregator to calculate the count per group. After the join filter on count <= 5.

Welcome aboard.

Your question is not really related to the topic of this thread. Would you please being a new thread, rather than hijack one that's already some years old?

... except where this convention might lead to duplicates, when you start to get ~2, ~3 and so on at the end of the basename.

Performance tip: the Char() function is quite expensive to evaluate. If you need them, initialize a stage variable so that they're only evaluated once.

Under the covers the Char() function is evaluated by reference to to a terminfo database.

Since you have to convert single- and double-quote characters you have two choices. You can use backslash as a quote character Convert(\@#$"'\, \ABCDE\ column) (I'm not sure if this works in parallel Transformer stage.) Or you can build the string using concatenation. Convert("@#$'" :...

Windows 7 is not a supported operating system for DataStage version 7. Anything you get to work is pure luck.

You need a separate (or at least a suitable) schema file for each source data file. The schema file name and the data file name can both be job parameters.

Null is internally represented identically in every row. Therefore every row for which the hash key column is null will go to the same partition.

WebSphere Information Services Director (WISD) is the latest name for what we used to know as RTI (real time interface).

Stay tuned for future changes.

That was a cynical statement that assumes that the vendor changes product names periodically whether they need to or not. It's a marketing thing.

A continuously-running job will never purge its log. At some point, therefore, the log will fill (reach 2GB if you're using the local repository) and your job will crash and burn. Have a design that runs - perhaps - for 23 hours 59 minutes and is scheduled to restart one second after that. Build in ...