DSXchange

SLEEP hh:mm

Not documented in DataStage BASIC manual.

Been around since the early days of UniVerse, but not documented in UniVerse BASIC manual either.

One of those undocumented features no-one seems to have a list of.

Yes, it's called Linux. Same hardware, sparse lookups are supported. I don't believe there's much you can do at the software level - convince "them" that more memory is required. And more disk, because you may need to spill into the scratch disk areas. Is it possible to perform a left outer join in ...

Heck, yes. A Filter stage or a Transformer stage with 100 output links. Easy.

Sounds like you're looking for the External Filter stage or the Wrapped stage. The former you will find in the Parallel Job Developer's Guide while you will find the latter in the Parallel Job Advanced Devloper's Guide.

Well, you can always work at the millisecond level, which IS supported. Indeed you can get the values for CPU and elapsed time (to millisecond accuracy, rounded) in log messages, so you don't even need to calculate them yourself.

Was the second post a "jitter on your Windows box"?

Paths. Command directory search path (PATH) and shared library directory search path (LD_LIBRARY_PATH or SHLIB_PATH or LIBPATH depending on your operating system).

Not NAP for sleeping up to 30 minutes!

It is really bad practice to call DSGetLinkInfo() from within a Transformer stage, because you have to call it for every row processed, and the answer is already there in @INROWNUM or @OUTROWNUM system variables. The overhead of calling it repetitively and unnecessarily will cause poor throughput, a...

LOOK at the error message! What you say is not what you did, or the error message is not being generated from the Hashed File stage you have been tampering with. You claim to have specified C:\HashFiles\Hash_File2 as the directory pathname. Why, then is DataStage trying to create a hashed file calle...

Quite.

Create a routine. Use the fact that the SLEEP statement has two syntaxes - one that sleeps for a given number of seconds, the other that sleeps until a particular clock time. Being in a routine, it is easy to calculate the next 30 minute point.

Welcome aboard. :D

You can reset parameter defaults in the Director client, but this does not change the design-time defaults.

Contact your support provider to ask whether there are any known issues in your version of DataStage related to what you are trying to accomplish.

Import errors are row #0 are typically caused by neglecting to set the First Line is Column Names property to True.

The raw form of the "data" - here { I D } - helps to confirm that this is the case. Clearly "ID" can not be an integer.

It is not possible to add 1.5 days to a date. It is, however, possible to add 1.5 days (36 hours) to a timestamp. There is no AddDate() function out of the box, though you could write one if desired. Otherwise, use one of the approaches that kumar_s suggests. Note that there are 86400 seconds in one...

This is a property of the Oracle bulk loader stage (or the Oracle Enterprise stage). The answer, therefore, is "wherever you specify". Right now I don't have access to Enterprise Edition, so can not research what the default value for this property is.