Search found 53125 matches

by ray.wurlod
Tue Mar 19, 2013 10:41 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Insert records from where it last failed
Replies: 7
Views: 2311

No, because that's not how Round Robin partitioning works. In any case, what happens if the failure occurs half way through a Data Set?
by ray.wurlod
Tue Mar 19, 2013 6:27 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: SCD Stage Operators
Replies: 7
Views: 2110

Range lookups also compile to transform operators.

I'd suggest that you need to get a compiler licence.
by ray.wurlod
Tue Mar 19, 2013 6:26 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Insert records from where it last failed
Replies: 7
Views: 2311

Welcome aboard.

The short answer is "No".

You have to design in a mechanism to "remember" which rows were inserted if you don't want to query the target.
by ray.wurlod
Tue Mar 19, 2013 3:00 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: transformation
Replies: 15
Views: 3877

True. But the solution is still straightforward using DataStage BASIC (for example in the BASIC Transformer stage). Convert the words into a dynamic array, take the leftmost character of each element, then delete the dynamic array delimiters. Convert(@FM, "", Substrings(Convert(" &quo...
by ray.wurlod
Tue Mar 19, 2013 2:57 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Issue with config Oracle Connector
Replies: 1
Views: 1504

Does your shared library search list include all the parent directories of the shared libraries mentioned? Are the permissions on these libraries set correctly?
by ray.wurlod
Tue Mar 19, 2013 2:54 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: SCD Stage Operators
Replies: 7
Views: 2110

The easiest way to find out is to look at the score for that job.

Then, maybe, you can report your findings back to us.
by ray.wurlod
Tue Mar 19, 2013 2:53 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: transformation
Replies: 15
Views: 3877

CAPITALS is a Transform that encapsulates Oconv(%Arg1%, "MCT")
by ray.wurlod
Tue Mar 19, 2013 1:27 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: transformation
Replies: 15
Views: 3877

Use BASIC Transformer stage and Oconv(InLink.TheString, "MCT")
by ray.wurlod
Tue Mar 19, 2013 1:26 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Range Lookup for the multiple reference dataset
Replies: 1
Views: 909

Using three Lookup stages is the correct approach, not least because it works.

When you use a range lookup the Lookup stage compiles to a transform operator and, as you noticed, gets finicky about what's able to be used (re-used) as a lookup key.
by ray.wurlod
Tue Mar 19, 2013 1:22 pm
Forum: General
Topic: After Migration Iconv & Oconv have different results.
Replies: 6
Views: 1720

Or you could simply change the NLS locale, particularly the TIME category.
by ray.wurlod
Tue Mar 19, 2013 1:21 pm
Forum: General
Topic: Resolve local host datastage 9.1
Replies: 1
Views: 2801

You need to test (for example with dnsping or similar) that the name can be resolved using DNS. If it can't a temporary fix is to put the actual hostname and IP address into /etc/hosts file.
by ray.wurlod
Tue Mar 19, 2013 1:18 pm
Forum: General
Topic: Unlocking the datastage jobs
Replies: 16
Views: 7028

mikva2012 wrote:what will be the impact if will run the cleanup_abandoned_locks.sh script from Unix command line.
Abandoned locks will be cleaned up. However, a lock held by another DataStage client is not an abandoned lock.
by ray.wurlod
Tue Mar 19, 2013 1:13 pm
Forum: General
Topic: Updating a Sequential File
Replies: 5
Views: 837

I do this a lot when updating values files of Parameter Sets. One approach is to overwrite the file entirely, using a Sequential File stage. That's the way I normally do it, particularly when the number of parameters in the Parameter Set is small. The other is to use a Hashed File stage in a server ...
by ray.wurlod
Tue Mar 19, 2013 1:10 pm
Forum: General
Topic: DataStage Phantom Aborting with @ABORT.CODE = 1
Replies: 3
Views: 2008

@ABORT.CODE is only set by server/sequence jobs. However, this is the final message. You need to show us earlier errors from the job log, which is where the cause is reported. These errors might even be reported in the log of a job called from a Job activity in your sequence.
by ray.wurlod
Tue Mar 19, 2013 2:42 am
Forum: General
Topic: Unlocking the datastage jobs
Replies: 16
Views: 7028

If you have IS Administrator role, log in to the Web Console for Information Server, open the list of active sessions and disconnect the session that owns the log (the session ID is in the "locked" message). The lock will be cleared immediately.