DSXchange

...We release the DS Job which embeds the code in the routines used within the job... Routine object code is not embedded within the job, it is stored in a different location (DS_ROUTINES and a pointer in the VOC). If the routines are transported as part of a release they are put in this location. ...

Indrani,

can you db2connect to the database from the DataStage server?

Pilot, the designer of the PX null handing system (an illegitimate issue from Mary Shelley and Freddy Kruger) has come up with some really wonderful ideas for torture - and when combined with the PX Sequential file stage (coded by a temporary resident from the depths of the 7th level of hell who was...

Which database are you connecting to? If it is Oracle, can you use tnsping on your DataStage system?

50 stages is a lot of steps. If you have an 8 node configuration then you are kicking off 400+ processes for this. Many PX sites might not have a single job with that many stages, but fire off several smaller jobs in parallel that start up thousands of pids. A lot of these PX processes will not do m...

Jinm,

theere is no routine code embedded in compiled jobs, all that they contain are links to the routine, and these are snapped at runtime with whatever the current routine object code is.

There are really just 2 attributes to configure in this file, firstly whether or not you want the deadlock daemon to run and secondly what the interval should be in seconds, defaulting to 15 minutes.

Ken, thanks for the help. The aggregator worked as you stated, but in this case the link collector and a subsequent transform with two stage variables will work for me.

Ken, I understand what you meant now - I orginaly assumed that it would do a complete data sort {just as that durn merge stage is obviously doing}. But do you think that the round-robin collector could ever get out of synch? I wrote small test job and started it 50 times (thank god for multi-instanc...

Jinm, as Roy has pointed out, the jobs that call a routine use only a pointer to it that is resolved at runtime, so you only need to re-compile the routine and any jobs that use it will get the new version; any jobs that are actively running using this routine will continue to use the old copy until...

Ken, since these data streams are going to contain hundreds of millions of records I really don't want to do any sorting at all. In reality I don't have two files, I actually have just one stream that has been split in a transform, with one stream going into a very complex shared container which ret...

jinm, I've made my own Japanese maps and manually added all the gaiji characters before, and the effort was all manual within DataStage/Universe. The tools might be unwieldy and archaic but they do work quite well - but only once you've gone through the painful process of using them. I'm curious as ...

Actually, you can have multiple instances running on the same machine at the same time. Plugins are not an issue with colocated versions 7.1 and 7.5; and I am fairly confient that it would work with 5.x and 7.5 - but not enough to recommend trying it on a production environment. I think you ought to...

Ken, thanks! Right after I posted the query I thought about the link collector, and wrote a test job using that. Essentially the rows alternate between File A and File B so with a stage variable in a transform it works perfectly. The speed is about 125,000 rows per second on this box (I'm using name...

I have two sequential files, each with two columns - the first column is the 'key' and the other is data. The source files are in the same order and of the same length - meaning that the n-th row of file A will have the same 'key' as the n-th row of file B. I want to output three columns, the key, F...