Search found 53125 matches

by ray.wurlod
Sun Dec 17, 2006 2:39 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Sequencer Questions
Replies: 18
Views: 5633

DSguru2B wrote:I agree. Thats why i never use UserStatus. I like things being physically written to the disk. Especially if its controlling the process.

As noted, user status IS written to disk. That's why it's accessible long after the job finishes. It's stored in the RT_STATUSnnn table for the job.
by ray.wurlod
Sun Dec 17, 2006 2:37 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: regarding pivot stage
Replies: 3
Views: 1140

Firstly, your "and vice versa" is incorrect. The Pivot stage only performs horizontal pivoting (columns to rows). It does not perform vertical pivots. The manual ( pivot.pdf ) gives a number of examples, and explains precisely how horizontal pivots work. I see no need to recapitulate that material h...
by ray.wurlod
Sun Dec 17, 2006 2:32 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: regarding date transformtions
Replies: 19
Views: 4750

Oconv(Iconv(Field(InLink.TheTimestamp, " ", 1, 1), "DYMD"), "D/DMY[4,2,2]") : " " : Field(InLink.TheTimestamp, " ", 2, 1) Don't use DATE.TAG data element. A Timestamp is not a DATE.TAG. Use Default or Timestamp if you...
by ray.wurlod
Sun Dec 17, 2006 2:30 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Comparing CFF files , Record to Record !
Replies: 11
Views: 3068

In the Transformer stage you open the stage properties window (use the leftmost tool on the toolbar), then select the Variables tab. You will need one stage variable for each field. The result of the expression is whatever you like, either a true/false, or a "flag" as suggested earlier. You may also...
by ray.wurlod
Sun Dec 17, 2006 2:27 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Folder stage to xml stage
Replies: 1
Views: 677

Failure to set NLS map may indicate that the executing user ID lacks write permission to the /invoice/NOV_Invoice directory. Setting an NLS map involves writing a hidden file called .nlsmap into the directory.
by ray.wurlod
Sun Dec 17, 2006 2:25 pm
Forum: General
Topic: odbc insert
Replies: 9
Views: 3991

Within a Transformer stage, use a NULL-handling expression in each case - maybe create a NullToValue Transform. But, as Craig suggested, you can do this within the extraction SQL. Depending on the database the function may be called something like NVL(). COALESCE() selects the earliest non-null valu...
by ray.wurlod
Sun Dec 17, 2006 3:50 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Folder stage to xml stage
Replies: 1
Views: 746

Failure to set NLS map may indicate that the executing user ID lacks write permission to the /invoice/NOV_Invoice directory. Setting an NLS map involves writing a hidden file called .nlsmap into the directory.
by ray.wurlod
Sun Dec 17, 2006 3:47 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: regarding date transformtions
Replies: 19
Views: 4750

Whenever an expression is red, right click on it and choose Validate Derivation - let the expression editor tell you what is syntactically invalid.

The original expression (as you posted it) lacks a final right parenthesis.
by ray.wurlod
Sat Dec 16, 2006 6:48 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: DS Error
Replies: 4
Views: 1338

And, in future, please try to come up with a more meaningful subject for your posts.

This one falls into the "my car won't go, what's wrong with it" category.
by ray.wurlod
Sat Dec 16, 2006 6:47 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: Error: code=-99 [General repository interface 'other error']
Replies: 7
Views: 4213

When you get a BLINK (backward link) error at this address it typically indicates an attempt to take a hashed file beyond 2GB, which is not possible with 32-bit pointers (and which, therefore, is why the structure becomes corrupted). You must monitor your hashed files and convert them to 64-bit addr...
by ray.wurlod
Sat Dec 16, 2006 6:44 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: how to count the number of delimiters in a reord
Replies: 14
Views: 9213

Index() is the wrong function. It reports the character position in the string of the nth occurrence of the substring.

Count() or DCount() is the correct function for counting instances, or fields delimited by instances of a delimiter, respectively.
by ray.wurlod
Sat Dec 16, 2006 6:42 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: creating users
Replies: 5
Views: 1502

Is it Windows XP Professional? Windows XP Home simply does not work. Second, do you have a network card active - that is, receiving an IP address from anything that can serve one (even a router)? You may find this to be necessary in the DataStage-on-XP environment. Third, have you rationalized domai...
by ray.wurlod
Sat Dec 16, 2006 4:53 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Single Large Project Versus Multiple Small Projects
Replies: 6
Views: 1488

Not if the reusable components are adequately managed and documented (for example on an intranet). The built-in reusable components (those shipped with the product) are available in every project. So you only need to manage and document the in-house-developed ones. If you're using a source code cont...
by ray.wurlod
Sat Dec 16, 2006 4:51 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Sequencer Questions
Replies: 18
Views: 5633

You're both right. User status IS job-specific. What you retrieve is qualified with the job activity name (e.g. JobActivityName.$UserStatus) so that there is no ambiguity. If you always want the user status of job #1 even on a restart, that's OK. The user status is persistent, and can be retrieved a...
by ray.wurlod
Sat Dec 16, 2006 4:45 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Accessing Universe tables from one project to another
Replies: 5
Views: 2438

If you use DataStage/SQL, you can qualify the table name with the schema name. For example, in Project1 you can (assuming you have Project2 and the table EXAMPLE1 exists there) SELECT * FROM Project2.EXAMPLE1; Therefore, to detect differences, you might left outer join DS_JOBS from Project1 with DS_...