DSXchange

Is this a server job question (you posted in the server edition forum) or a parallel job question (you marked the job type as parallel)? Merge stage in the two different editions performs radically different tasks. Why are you insisting particularly on the Merge stage anyway? It's almost certainly n...

Most lookups performed by a Lookup stage are "in-memory" lookups, in that the reference table is pre-loaded into a virtual Data Set. This is manifested as a composite operator consisting of LUT_CreateOp and LUT_ProcessOp operators; the first loads the table, the second uses it to perform lookups. Yo...

All you need is no Record Delimiter property, and Record Length = Fixed on the Format tab (and, by implication, First Line is Column Names = False on the Properties tab). When importing the table definition for the Sequential File, you specify that it is fixed width format, and the width of each col...

Only with a custom routine or Build stage. The algorithm (psuedocode) would look something like:

The BASIC Matches operator is not being used for pattern matching in your example. Even in a server job I would not have used it. Your logic is more efficiently satisfied (in either environment, server or parallel) by: If vProdCode ="B" Or vProdCode = "C" Then "1100" El...

Welcome aboard. Mutex locks are one way that semaphores are implemented. You can search the forum for examples. You don't actually wait on a mutex lock - you "spin" (keep re-trying until you obtain the lock). Whatever has happened has taken too long; more time than is permitted. It could be a wait o...

You can most easily accomplish this with a job control routine. Where do the twelve sets of parameter values come from? The following example assumes one set of parameter values per line in an input file. $IFNDEF JOBCONTROL.H $INCLUDE DSINCLUDE JOBCONTROL.H $ENDIF $INCLUDE UNIVERSE.INCLUDE FILEINFO....

Or just move the substrings around.

No. Each job log is a separate database table in the Repository.

I guess you *could* theoretically create a UNION ALL of all those tables, but I suspect you would hit some command line length limit for a sufficiently large number of jobs.

No. You need one Trim() function per character.

You may like to Search the forum for a parallel version of the Ereplace() function.

"Free". Is this true in version 7, or only in version 8?

When importing an Oracle table definition prefer the Orchestrate ( orchdbutil ) method. This is known to import the metadata, particularly for decimal numbers, more accurately than other methods. There is no published tool, either IBM or third party, for importing a table definition from DDL. This i...

It's not a 2GB problem; otherwise the BLINK address would have been 0x80000000. The file may be repairable using fixtool; again, run this under advice from your support provider. Take a backup copy of DS_JOBOBJECTS before attempting any form of repair. Even the repair tool may have to discard some d...

You might try the output link variable LASTERR.

Otherwise, create a routine (possibly an after-job routine) to read the dbase log and write the message into the DataStage log. It ought not to be too difficult.

Column values only exist while a row is being processed. They are not stored anywhere within DataStage. Therefore the answer to your original question is that it is not possible to "get" the values other than within the job design itself, as others have noted.