Search found 53125 matches

by ray.wurlod
Tue Feb 19, 2008 5:51 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Parallel routine issue when using library file
Replies: 3
Views: 1603

You probably need an extra -l option in APT_LINKER_OPTIONS to refer to your shared object.
by ray.wurlod
Tue Feb 19, 2008 5:50 am
Forum: General
Topic: Exporting jobs through Commad prompt
Replies: 7
Views: 4406

You could investigate dsexport.exe command.

There is nothing on the server.
by ray.wurlod
Tue Feb 19, 2008 5:49 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Error in compiling the job
Replies: 6
Views: 1409

Post the exact message. The variable is named. It may be a stage variable, it may be a column name. It may be a redundant reference to a column in a constraint expression or in a stage variable initialization expression. Find it (the variable name) in your job design.
by ray.wurlod
Tue Feb 19, 2008 5:46 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Renaming a link breaks a transformer
Replies: 7
Views: 1901

Are you referring to the Key Expression on second (and subsequent) reference input links, or subsequent fields on the (only) reference input link? The latter do not need to change; they are not qualified by link name on the reference input link itself. However, some of your stage variable, constrain...
by ray.wurlod
Tue Feb 19, 2008 5:42 am
Forum: General
Topic: Importing dsx file server side (from a ksh script)
Replies: 2
Views: 1553

Welcome aboard. You have explored the entire range of tools - there is no general server-side import tool. The dsjob command, as you have noted, can only import job executables. The rest must be done from the client.
by ray.wurlod
Tue Feb 19, 2008 3:37 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: retrieving job status from dsjob
Replies: 8
Views: 2944

When you originally specified dsjob we understood that you meant to do this from the command line. Oconv() is not available at the command line. What you have done is appropriate within a DataStage job but, if you're going to do that, why incur the overhead of invoking dsjob command? There are perfe...
by ray.wurlod
Tue Feb 19, 2008 3:35 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: eliminate warning messages in Datastage server edition
Replies: 4
Views: 1608

Thorough analysis, cogent design, comprehensive implementation. Identify all possible causes of warnings in the development phase (unit testing) and code to prevent their occurrence. Then, in production, any warnings that do occur are not the fault of the ETL process, and may be regarded as genuine ...
by ray.wurlod
Tue Feb 19, 2008 3:32 am
Forum: General
Topic: returning failed record and feild
Replies: 5
Views: 1513

The log will contain any error message generated by the database server. This may or may not contain the name of a column. It may, for example, contain the name of the constraint that was violated. It's then up to you to examing the column values (also logged) to determine which value(s) caused the ...
by ray.wurlod
Tue Feb 19, 2008 2:39 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Lots of warning because of transformation
Replies: 4
Views: 1438

ByteLen() is available. It may stay red (because it is not in DSParams file) but it will compile and it will run.
by ray.wurlod
Tue Feb 19, 2008 2:38 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Conversion code
Replies: 9
Views: 2194

If Left(InLink.TheString,1) = '"' And Right(InLink.TheString,1) = '"' Then InLink.TheString[2,Len(InLink.TheString)-2] Else InLink.TheString You may choose to strip leading and trailing white space first, perhaps in a stage variable, if there is white space outside...
by ray.wurlod
Tue Feb 19, 2008 2:35 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Aggregator
Replies: 1
Views: 677

It is possible. Whether you can write the code to execute faster than that written by the experts depends on how good your programming skills are, and whether you can avoid some of the compromises they necessarily must take (for example you might only handle particular data types; they must accommod...
by ray.wurlod
Tue Feb 19, 2008 2:34 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Parallel job reports failure (code 139)
Replies: 9
Views: 6082

Sounds like you need to get your Oracle DBA on board, and maybe your network administrator as well, to find out whether it's Oracle that's raising the SIGPIPE signals and, if so, for what reason (for example unexpected delays). That Oshexecuter.sh has core dumped may indicate that something at the D...
by ray.wurlod
Tue Feb 19, 2008 1:53 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: SH: sqlplus: not found.
Replies: 6
Views: 3750

What is different between the good and bad runs? Are there any before-job subroutines or different paths through job sequences involved where the environment settings might be changed? Look at the log event called "Environment variable settings..." for each kind of job. Try to elucidate where they m...
by ray.wurlod
Tue Feb 19, 2008 1:51 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Performance of Aggregator
Replies: 2
Views: 960

It depends (as always). If the grouping columns are supported by indexes, then it may well be faster to do it in the database. If you have sorted data, then it may be faster to use an Aggregator stage. Is the database local or remote? If remote, it may be better to aggregate in the database, because...
by ray.wurlod
Tue Feb 19, 2008 1:49 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Parallel job reports failure (code 139)
Replies: 9
Views: 6082

This is some new meaning of "superior", then?
:lol:

Do you know what this port number is used for?