DSXchange

I already did. For less than 30c per day you can purchase premium membership which, among other things, allows you to read the entire posts of the five Premium Posters. All the revenue from premium memberships is devoted to the hosting and bandwidth costs incurred by DSXchange so that, by signing up...

What output is needed where there are case differences, such as "Ram sita ravan ram lakshman"? What output is needed where there are embedded substrings, such as "ram sita ravan ram lakshram"? I agree that this falls into the do-it-yourself category but, before you do, tighten your specification as ...

The only officially supported compiler is the one that comes with Visual Studio.

Look in /etc/services file.

Try removing the space character that precedes the ":" that introduces the new data type in each of the specifications. Make sure that symbol, Units and U_type are names of actual columns on the input link. Does u_typ exist in another specification (perhaps a KEEP or DROP specification)? Make sure t...

Operator combination is thwarting the diagnostic process - we only see that the errors are coming from APT_CombinedOperatorController. Try running the job with operator combination disabled so you can see which stage (operator) is actually generating the errors. At first glance it appears to be a pr...

I normally use a shared container to receive the error code, decode it, add the job id, run time, user, etc., and insert into the error table. You can have as many references to this shared container in your job as needed.

cd to the directory where you first executed the uvsh command and there execute the following UNIX command rm -rf VOC D_VOC VOCLIB D_VOCLIB \&SAVEDLISTS\& D_\&SAVEDLISTS\& This will clean up the unneeded account. DO NOT - ever - execute these commands in the DataStage engine director...

These documents are available from IBM (web site) and/or your support provider. Ask for "what's new in 7.5".

Has the job been deleted? Can you find it in any of the clients? Try re-compiling the job (which may not work). Try running the DS.REINDEX ALL command when you have exclusive access to the project - it may be that the CATEGORY index on DS_JOBS and/or DS_JOBOBJECTS needs to be rebuilt.

In the Routines branch of your Repository create a new routine and ensure that its type is Before/After Subroutine.

Within the subroutine, code in DataStage BASIC and use the BASIC SQL Client Interface, or BCI, functions to mimic operation of the ODBC API.

Search the forum for examples.

That is an insane requirement that you should push back against. What's wrong with using more than one stage? However, you could: create a plug-in stage (which would have to be written in C) rewrite the logic as a stored procedure and call that from an ODBC stage to deliver the rows into DataStage d...

... or get premium membership to see solution already provided.

Pre-load the lookup information into a hashed file. Only load the rows and columns that will actually be needed. Cache the hashed file in memory. Create separate links for insert-only and update-only operations. Consider using bulk loading for the insert-only stream. Purchase a more powerful server....

Within the after-job subroutine you can use DSGetParamInfo() function to obtain the value of each parameter. Use DSJ.ME as the job handle argument.