Search found 53125 matches

by ray.wurlod
Wed Nov 25, 2009 11:55 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Aggregator group change
Replies: 9
Views: 2485

(Auto) gives Entire as the partitioning algorithm on the reference input link to a Lookup stage. On an SMP environment, Entire is implemented by placing a single copy of the data set into shared memory. Depending on the circumstances, this can be seen with other complex stage types such as Transform...
by ray.wurlod
Wed Nov 25, 2009 11:51 pm
Forum: General
Topic: WARNING:Implicit conversion from source type to result type
Replies: 12
Views: 15368

Be consistent with your use of the extended property Unicode on your strings.
by ray.wurlod
Wed Nov 25, 2009 10:09 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Aggregator group change
Replies: 9
Views: 2485

Yes, provided that you do NOT check the "perform sort" check box on the Partitioning tab. That would force an unnecessary re-sort. The upstream Sort stage, being directly on the input link to the Aggregator stage, suppresses the insertion of an unqualified tsort operator. You will still se...
by ray.wurlod
Wed Nov 25, 2009 10:06 pm
Forum: Information Analyzer (formerly ProfileStage)
Topic: First Column Analysis results in error
Replies: 4
Views: 4263

Yes, I thought that there may have been a log stored in RT_SC32. Obviously not. Need to think about it for a while.
by ray.wurlod
Wed Nov 25, 2009 9:55 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Aggregator group change
Replies: 9
Views: 2485

No. Use a Sort stage.

Remember from the transition class - one task, one stage.
by ray.wurlod
Wed Nov 25, 2009 9:54 pm
Forum: General
Topic: istool syntex -Need to include all the items for the export
Replies: 2
Views: 1757

Read the documentation on istool. What is it meant to "export" into a deployment package?
by ray.wurlod
Wed Nov 25, 2009 7:13 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Aggregator group change
Replies: 9
Views: 2485

Output on a group change occurs if the Sort, rather than Hash, method is employed.
by ray.wurlod
Wed Nov 25, 2009 7:11 pm
Forum: Information Analyzer (formerly ProfileStage)
Topic: First Column Analysis results in error
Replies: 4
Views: 4263

Look in the RT_SC32 directory within the ANALYZERPROJECT project directory on the server.
by ray.wurlod
Wed Nov 25, 2009 12:21 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Warning Missing record delimiter "
Replies: 1
Views: 1874

The last line in the file is missing its line (record) terminator.
by ray.wurlod
Wed Nov 25, 2009 12:19 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: How to get rid of phantom process
Replies: 1
Views: 1601

dsjob -stop project jobname
by ray.wurlod
Wed Nov 25, 2009 12:17 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Need to trim Special Character
Replies: 7
Views: 2410

Were you to follow my advice you would mention, in the Trim() function, the stage variable whose value has been initialized to Char(9).
by ray.wurlod
Wed Nov 25, 2009 12:10 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Insert new rows(with the same key) breaks the run!
Replies: 7
Views: 9797

You can not bypass the check for existing row. Most efficient way is to load a hashed file with just the key values from the target table. If there are potential duplicates in your input stream, you also need to maintain the hashed file as rows flow through.
by ray.wurlod
Wed Nov 25, 2009 1:42 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Need to trim Special Character
Replies: 7
Views: 2410

Specify the tab character in the Trim() function as the character to be trimmed. The tab character can be represented as Char(9). Performance tip: Evaluate Char(9) when initializing a stage variable (perhaps called svTab). Then don't provide a derivation for this stage variable in the main part of t...
by ray.wurlod
Wed Nov 25, 2009 1:40 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Full outer join in Ds
Replies: 8
Views: 4916

A full outer join is not the same thing as a Cartesian product. Why not create an example with three rows on each input? Describe your expected output, and what you actually get. Then describe what you're doing on the Mapping tab of the Output link, which is where I think you're missing picking up i...