DSXchange

No, I don't think so. Most of the information is in the key itself (@ID) - splitting it out into separate fields was probably the whim of some programmer within the engineering group who look after this stuff.

Please post your uvodbc.config entry for this DSN. In particular I would like to know whether you have made any attempt to tune the throughput of this (or every) DSN using MAXFETCHBUFF and/or MAXFETCHCOLS tuneable parameters.

You could, of course, use DataStage to populate one data-only worksheet, and have formulae referring to these data cells on other worksheets in the same workbook.

You could, of course, use DataStage to populate one data-only worksheet, and have formulae referring to these data cells on other worksheets in the same workbook.

It's to be found in a subdirectory in your project (on the server machine). The name of the subdirectory depends upon the job number which, you will possibly be aware, is mapped in a table called DS_JOBS - you can search the forum for details. The name of the subdirectory that contains the .trx file...

Set APT_DUMP_SCORE to enable dump of the score, then examine the score in the job log. There you will see that DataStage has inserted a tsort operator on each of the input links to your Join stage. It has probably also inserted buffer operators, but that's a different topic.

Misunderstood. You can do it with stage variables in a Transformer stage, constraining the output for an exact match with what's in the buffer. You could probably leverage the Remove Duplicates stage.

INSERT INTO DICT DS_LICENSE(ID,CODE,LOC,NAME,FORMAT,SM) VALUES('LICENSE','D','1','License No','11L','S'); INSERT INTO DICT DS_LICENSE(ID,CODE,LOC,NAME,FORMAT,SM) VALUES('EXPIRY','D','3','Expiry Date','10R','S'); INSERT INTO DICT DS_LICENSE(ID,CODE,LOC,NAME,FORMAT...

You've been going about it the wrong way from the outset. You don't delete anything at the operating system level - if you do, the "delete file before create" fails - it expects to find all three components (the VOC entry as well as HHH and D_HHH. Now you need to find a "clean point". Try using a di...

The C++ code for a parallel Ereplace() function has already been posted in this forum. Search for it.

Instead of using job parameters you could use stage variables. Initialize them, and do not derive them further - that is, leave the derivation field empty.

subargs={asc} means that a sort was specified to be in ascending order.

Not enough information. We have no idea what these jobs are doing, and therefore no idea why the second job may be waiting. For example it may be locks or semaphores within DataStage itself (that you have elected to use to synchronize events), it may be locks in a database, it may be other things. T...

APT is an antonym (the antonym?) for INEPT.

What else is happening on the network?

The worst case of this symptom I ever encountered was during a Cricket World Cup (I was in India). It turned out that 75 developers (and probably lots of others) were watching the game live in their browsers. Goodbye bandwidth!