DSXchange

A Data Set is the fundamental unit of data container in DataStage parallel jobs. A Data Set is the only structure which Orchestrate operators - and that effectively means stages in DataStage jobs - can deal. Every link in every parallel job (unless optimized out) has a virtual Data Set associated wi...

DSGetLogEntry is the one you want.

There is no direct API equivalent to dsjob -logdetail command, which makes multiple DSGetLogEntry calls when it executes.

Is this table created by a Hashed File stage, a UniVerse stage, a CREATE.FILE statement in TCL, a CREATE TABLE statement in TCL or a mkdbfile command? Check that there is no non-printing character as part of the key value. The id column has an inferred data type of Integer, but a non-printing charac...

St Jude were there, and even presented. So at least some of them are probably up to speed.

I think, for whatever reason, everyone comes away from the conference exhausted. It'll take a couple of days for everything to settle.

For individual jobs you would need to change the //PURGE.SETTINGS control record in the RT_LOGnnn hashed file. This can be done with SQL but there are a couple of tricks required around data types. And I can't guarantee that days > 99 is supported (but can't think of a good reason why not).

Ah.

The GUI only allows 99 (the spin control is limited to two digits).

More would not really be sensible, but might be possible by editing the auto purge setting the the DSParams file.

Doesn't just CurrentDate() alone do it? If not, convert its data type to string.

Attended an IOD 2009 session on capacity planning for Information Analyzer last night, some of the numbers are just plain scary! In terms of DataStage, it really does matter how you organize the tiers. In development it's a good plan to separate the DataStage engine from the WAS and repository tier,...

Welcome aboard. Numeric(38,2) can contain at most 36 digits to the left of the decimal point. You might use the Num() function.

What I posted. What was incorrect were your two "unlike" phrases.

You don't need to load it into a hashed file at all, unless you're interested in best performance. Both inputs to the Transformer stage can connect to database stage types. However, do consider shifting the join to the database if possible (the database server is not overloaded). That way DataStage ...

Make the record delimiter "DOS format" might work. Or are the 0D0A pairs within quoted strings? Otherwise, to convert the NUL characters to space, define a stage variable svNUL initialized to Char(0) then: Convert(svNUL, " ", InLink.TheColumn) To replace the CRLF pairs, define a ...

There's another cool tool called (tentatively) Watchdog for InfoSphere that can set up and monitor events of various kinds all over the InfoSphere range of products. This one is still very much in the requirements gathering phase (one good reason to come to IOD is to have a play with, and provide in...

Ordinarily that would mean that MESSAGE_NUM is not a valid column name in the table that you're loading. Do check that you specified the correct table name.