DSXchange

The system variable @DATE or the macro DSJobStartTimeStamp can be used. These are set when the job sequence starts, and do not change thereafter.

Since you're on version 5, there are other techniques available in later versions - such as user variables - that aren't available to you.

Looks like it's getting through to UniData OK (since the message "You have no privilege on file PERSON" is being returned from UniData). Check the privileges you have on the various UniData files with which you seek to work.

The numeric part of stage and link names purely reflects the order in which they are added to the design canvas. The format of this number is embedded within the DataStage source code (which you don't have, and is therefore not negotiable). As for default database and table name, the only sensible d...

That you can import table definitions and view data tells you that the DSN is properly configured. The question then is why SQL Server is refusing to deliver - is it the second row of the query? Does some external activity prevent the cursor from being advanced? Can you run the SELECT statement used...

An approach using stage variables may be found here

Without seeing the export file, I'd hazard the guess that the export file has somehow become corrupted, probably in the J\60\V0S0 record, which will contain a list of pins connected to that stage. The only fix would be to export again. Can you describe what the exported job looks like?

Other than get a user ID that doesn't include a "$" character, no.

Are you correctly identifying the host name when trying to connect from work? Try using the IP address rather than the machine name, to eliminate the possibility that gethostbyname() is the culprit.

If you've already got months and counter, then assuming months + counter can never be less than 1 or more than 24. All data types are uint8 or a larger or signed integer.

The Lookup stage can perform conditional lookups.

If it's Type 2 SCD there is no such thing as "update the previous record". The old record is expired as soon as any update comes through. In that way history can be re-built as required, which is the point of Type 2 SCD. The remainder of your question is therefore moot.

scorpion wrote:can any one please respond !

Some of us sleep occasionally. You waited only one hour before posting the above. If you require urgent support, contact your support provider and sign up for a premium support program.

You can add more scratch disk resources to your configuration file. Each processing node can (should) have multiple disk and scratch disk resources (on different disks) so as not to create an I/O bottleneck.

Where an input row does not match the metadata in the schema for its link.

Basically, no. It's an iterative heuristic. You bucket them as either unhandled patterns or unhandled strings, then either add to the PAL for your rules or create rule overrides.