DSXchange

1. Only via uninstall and reinstall. But it can be disabled by setting NLSMODE to 0 in uvconfig and regenerating the image. 2. What recommendations? I haven't had any problems, but again I haven't tried to stress test it. 3. In Canada you won't get much benefit, maybe the proper handling of accented...

That's because it doesn't belong in VOC. VOC provides the vocabulary when you are working in a DataStage shell; kill is a command that is only issued in a UNIX shell. But, a major caveat, you use kill at your own peril. Be aware of the risks, be aware that parallel jobs have lots of child and even g...

kenstorm wrote:I am going crazy here.

You're earning your qualification (you have to be crazy to work here).

The only limit is any imposed by the operating system. QualityStage itself does not impose a size limit. As far as I am aware there is no mechanism for splitting the PAT file (which is an output from an Investigate stage) into more than one.

No you didn't. You got rid of it by supplying either a value or a job parameter reference to the missing UserID property.

If you think the issue is resolved, please mark it so, to assist future searches.

Alas, you can't use -nrecs 2.5 !

Sequential files have neither keys nor data types. You can identify columns based upon the metadata, but that's all. Use a profiling tool to determine whether there are any candidate key columns. A candidate key column contains no nulls and no duplicate values. Profiling tools such as ProfileStage, ...

You can extend the functionality of DataStage by creating new stages that invoke other operators, or operators that are used in existing stage types but in different ways. You can also use a Generic stage to invoke an Orchestrate operator directly, if you don't want to go to the trouble of creating ...

How about marking the topic as resolved?

It is a clean number. View Data always adds sufficient leading and trailing zeroes so that you can verify that precision and scale are being handled correctly. And View Data always displays a decimal placeholder when the data type is non-integer numeric.

Are you running on a two-node configuration? The -nrecs option is per node. See Orchestrate Operators manual, page 10-3 (version 7.5).

The performance is allegedly fantastic. You use a dynamic configuration file, and let the grid management software do its load balancing thing. Send an email to Bob Johnson at IBM - he's implemented a couple of them.

There's no such thing as a regular Data Set. They are all created within the Orchestrate framework. There is more functionality in Orchestrate operators than is accessible through DataStage stages, primarily through using different command line options. That is why the option exists to create a Cust...

Not legally.

This isn't a server job, is it?