DSXchange

You have Enterprise Edition of DataStage, and you probably work in parallel jobs, you use the Parallel Job Developers Guide to build those jobs. Ascential have cunningly hidden job control documentation in the Server Job Developers Guide, even though it pertains to all job types and not just server ...

It was such an intimidating subject heading I found it difficult to even open the thread, the full error message is even scarier. Perhaps you could soften the impact with something friendlier such as "an interesting parallel job error". Fill us in with some background, is this happening to all your ...

In a transformer the AlNum function tells you whether any non alpha numerics exist and the convert function can replace them. Convert supports a fromlist of strings and a tolist. You put all your no alphas into the fromlist and a space in the tolist. There are very few examples of parallel functions...

If you run the dsjob -report with report type set to BASIC you get Job start time, Job end time, Job elapsed time and Job status. It would be quite easy to read this file in a basic routine and write it out in the format required. For example FIELD(FIELD(InputRow, "=", 2), " ", 1) gives you the date...

Isn't that what copy and paste of a stage is for?

There is also a FAQ "Where do I get information on DataStage" that will lead you to various sites. The most reason release notes for DataStage can be found on devnet.

It is well documented in the archives of this site. Do a search for failover or recovery or banding to find threads on the subject.

The demand for DataStage will only keep going up as the legacy coders get older and the manual coding methods get less attractive. Soon it will be easier to find a DataStage person then a cobol code cutter or a PL/SQL developer forcing companies towards a tool based approach. DataStage Hawk release ...

The main decision is whether you keep it on the mainframe, via parallel jobs on mainframe Unix System Services, or move it entirely off the mainframe onto Unix or Windows servers. I have heard that parallel jobs on USS are extremely effective at gobbling up CPU and resources to run at very fast spee...

I haven't checked out the job sites in a while but a search of the Australian job site seek showed 110 DataStage jobs! I am used to seeing up to 20 jobs at any one time. Informatica had 49 job listings. There is a lot of duplication of listings on these sites and I figure if you divide it by three y...

Native Oracle CDC is available for an additional license fee and performs log scraping against the RDBMS. Haven't used it myself. Other CDC options involve standard techniques such as retrieving rows based on creation and update dates or retrieving all rows and filtering on the server based on chang...

Nice idea though. How are you delivering jobs to production? If you use export/import maybe you can put a job description annotation into all your jobs and when you deliver it to production do a search and replace on the export text file to replace the annotation colour property to give it a differe...

To avoid a transformer altogether you could try Column Import followed by a Modify stage, if the derivation can be shoe horned into a Modify conversion function, or a Column Generator stage if it can be derived in the column properties of a new generated column. If you have to have a transformer the...

Maybe \0x00 which is the character representation of a null. As with most of the DataStage documentation the Default option comes with almost no examples or syntax clues.

Immediately when I see input of 1,1,1,1,3,3,3,3,2,2,2,2,4,4,4 results in an output from the join of 1,3,4 I would look at sorting problems. The join stage needs both links to be sorted and usually adds a join if one does not exist. Check your symbols on your links to see if the sort appears on both ...