Search found 53125 matches

by ray.wurlod
Thu Jul 15, 2004 2:52 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Date values out of and into Microsoft Access
Replies: 11
Views: 3098

Possibly the indexing in the Repository needs re-building. Use the Cleanup button in the Administrator client, or the Re-Build Indexes option from the DS.TOOLS menu on the server.

Are you selecting "DS Routine" from the expression editor's operand menu?
by ray.wurlod
Thu Jul 15, 2004 2:50 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Balancing input and output record counts
Replies: 13
Views: 3611

Au contraire, it's extremely useful, as it includes templates for every piece of documentation (MS Word, MS Excel and MS Project) that you need. But it would have taken too long to explain all this! Get someone at Ascential to send you just the schematic (a single A3 sheet). If there's enough intere...
by ray.wurlod
Thu Jul 15, 2004 2:44 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: date function
Replies: 7
Views: 4063

That's why I suggested a "BASIC Transformer" stage, not a "Transformer" stage. :wink:
by ray.wurlod
Thu Jul 15, 2004 4:34 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Date values out of and into Microsoft Access
Replies: 11
Views: 3098

Here it is!

Code: Select all

The functions I can see in Manager's Routines\Examples\Functions: 
ConvertMonth
ConvertQuarter
ConvertTag
ConvertWeek
ConvertYear
QuarterTag
Timestamp     <<----------
WeekTag
by ray.wurlod
Thu Jul 15, 2004 4:28 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Removing quotes from strings
Replies: 8
Views: 2127

Keep in mind that there are three quote characters available in DataStage BASIC expressions. \The title is "Alice's Restaurant."\ is a valid string constant. If you're only converting on a character-by-character basis, Convert() is a more efficient function to use than Ereplace(). To remov...
by ray.wurlod
Thu Jul 15, 2004 4:24 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Balancing input and output record counts
Replies: 13
Views: 3611

Correctly the name is all upper case (ITERATIONS). Allegedly it's the methodology used by all Ascential consultants world-wide, but I have seen no evidence of that. It is a very comprehensive methodology; the first steps, for example, seek to determine whether the enterprise is even ready to adopt D...
by ray.wurlod
Thu Jul 15, 2004 4:18 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: date function
Replies: 7
Views: 4063

The end day of the previous month is the day before the first day of the current month. The latter is easy to construct. I will have to check what DateDiff functions are available in PX to do the "day before" part. If the data are coming out of a database, it's almost certain to have date manipulati...
by ray.wurlod
Wed Jul 14, 2004 2:53 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Exception Handler
Replies: 2
Views: 1041

Most activities support multiple outputs. For example, a Job Activity might have three outputs, triggered by "Run OK", "Run with Warnings" and "Otherwise". You can use Custom triggers anywhere. The Exception handler is a stand-alone component that is triggered if something goes wrong (read "aborts")...
by ray.wurlod
Wed Jul 14, 2004 2:51 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Date values out of and into Microsoft Access
Replies: 11
Views: 3098

MS Access timestamps include milliseconds, so anything you generate to perform comparisons also need milliseconds. The easiest way to do this is to concatenate ".000" to an ISO8601 timestamp.

Beware, too, of case sensitivity. "Timestamp" and "TimeStamp" are not the same thing.
by ray.wurlod
Wed Jul 14, 2004 2:46 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Abnormal Termination
Replies: 4
Views: 3665

There's no parameter. There IS, however, good design and good operational practice. One of these is determining the cause for the aborts (reset the job and read the log entry called "from previous run") then fixing the design so that they don't recur. Another is to limit the total number of simultan...
by ray.wurlod
Wed Jul 14, 2004 2:44 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Strange behaviour of uvodbc.config file
Replies: 3
Views: 2635

Any time you import an ODBC table definition DataStage "updates" - and therefore re-creates if necessary - the uvodbc.config file in the project.
There is no configurable option to suppress this behaviour. :cry:
by ray.wurlod
Tue Jul 13, 2004 2:47 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Global Routines across projects
Replies: 6
Views: 1730

It CAN be done; this is what is referred to by "custom UniVerse function" as a type of Routine. The Routine, in this case, is an interlude to the custom function, which may be either locally or globally catalogued.
by ray.wurlod
Tue Jul 13, 2004 1:57 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Date Format
Replies: 3
Views: 866

For single-character replacement, Convert is more efficient than Ereplace. The original YMD follows the default rule, returning the shortest possible string for each component. This is where the zero from the month vanished to. Specifying [4,2,2] forces the conversion to use that many digits for eac...
by ray.wurlod
Mon Jul 12, 2004 7:58 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Generate Multiple Rows from a single row
Replies: 3
Views: 1050

Are there always exactly three characters in Field002? If so, you can use something like this. source ----> Transformer -----> SeqFile -----> Target The Transformer stage writes a single column to the SeqFile stage, of the form Key:",":Field001:",":Field002[1,1]:NL:Key:",&qu...
by ray.wurlod
Mon Jul 12, 2004 7:54 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Sequential File: Too many columns recieved than expected.
Replies: 3
Views: 1095

You could re-design, where you read the entire record as a single column, then parse out only the pieces you require.
Using this design, you can also report on erroneous input data.