DSXchange

In DataStage you seldom have the luxury of being able to choose your source data or target systems. The question is quite irrelevant unless you have specific circumstances that you haven't mentioned. Flat Files - compact - can be given any name you wish and be put anywhere on your machine - can be w...

If you can sort your incoming data stream to the aggregator on the columns you use in that stage and then specify in the aggregator that the columns are sorted you will see that the stage performs blazingly fast. This is because it no longer needs to store data temporarily, it just sums/aggregates v...

DataStage itself would need to be reinstalled only when changing the machine name.

One or more temporary files are only created when needed, and this depends upon how you program your job(s).

What does this job do with SDKSEQUENCES? Does it use the routines? If so, then you probably deleted the UNIX file SDKSEQUENCES in your project directory without modifying the VOC. Can you check whether or not this file is present?

Congratulations on your success!

The pros and cons of using a sequential file can only be discussed when you have alternatives. Are you comparing using flat files to a database? Hashed Files? Printouts?

I'm not sure what you mean by "command level difference"; the two statements returns the number of records in the two different hashed files when called from the TCL shell. Could you explain a bit further as to what you are looking to answer?

A couple of things need to be modified in the previous comment: Types 2 through 18 are static hashed files each based on the hashing algorithm / nature of key column data While true, the important aspect of static hashed files is not the algorithm used to turn the key into a number (the hashing meth...

What have you set your commit frequency to and what does your DBA say is happening on the database when the "hang" condition occurs? How many nodes are in your APT_CONFIG file?

Could you post the error message text onto this thread? That is quite odd, particularly as normally the PX install also encompasses the server .

I missed the bulk load part, was just thinking in terms of a normal load. I think the bulk loader in Oracle might take an exclusive table lock as well, but can't recall exactly.

That specific error was the bane of my life at version 3 and 4 of DataStage and an older Oracle 7 years ago; it was always sporadic and very hard to trace. It was invariably an Oracle issue and not a DataStage one, through and I haven't seen that error crop up in years. As mentioned above it is impo...

There is nothing stopping you running multiple instances of your job (without the truncate being called by each instance). But the bottleneck might be the speed with which Oracle can load the data as a whole and not per process. I am pretty sure that just splitting up your load into separate streams...

Yes, server and parallel can and do coexist in the same project.

What did you install (did you add a license for Server?) and what exactly no longer works in parallel?