DSXchange

The Parallel stages will only show up when you have installed the PX package; so it would seem that the installation only did server. Did you check your license xml to see if you actually do have PX?

Are your APT_CONFIG files the same, or have the same number of nodes, in both installations? Often a job developed on a 1-node site will show errors when run with more nodes (the opposite is not true, which is why jobs should always be developed on a 2 or more node configuration).

Vincent - you were so much more direct than I - do you think I was being too subtle regarding a 1-node configuration :lol: Compare my 1-node configuration doesn't give PX a chance to take advantage with your somewhat more forceful A one node DataStage configuration for a Teradata environment is insa...

In order to do this I would recommend not looking into the undocumented DS_JOBOBJECTS hashed file, but using the documented API to get this information. Among other calls, DSGetStageTypes will return all active and passive stage types used in a job. If you really feel the need to explore in the repo...

You can create a server job without inputs. Create a transform stage with no inputs, define a stage var (you don't need to use it) and define a constraint such as "@OUTROWNUM<2" and you are done.

Not 8.1 but 8.x is the dividing line.

At versions of DataStage prior to V8 the metadata repository was within the tool itself, stored in hashed files. As of Version 8 this changed and there is now a requirement to use an external repository. DB2 is delivered as part of the package for use as a repository but you can also use Oracle. Use...

I posted some code for a user-written function that converts integer values to COMP-3 here on DSXChange a couple of days ago, if you use the search function you can find it.

Let us take this step-by-step, starting with the source. The COBOL data picture of " S9(16)V99" is not a packed field. It would be packed if the definition stated " S9(16)V99 USAGE IS COMP-3" or similar. Next, the target column datatype of DECIMAL is an SQL data type, and unknown...

Statistics in the monitor may be inaccurate during the run, but upon successful completion of the job will be accurate. I am not sure if, after your "force compile" the job is still executing differently (i.e. incorrectly) from your copy of the job, which you stated runs correclty. Try run...

When you run your job is the CPU use at over 90% for long periods of time, or is I/O overloaded or does the machine swap? Without knowing where your bottleneck is one cannot recommend tuning.
Although a 1-node configuration doesn't give PX a chance to take advantage of any parallelism whatsoever.

Why did you once again quote your message AND add nothing to the thread but just repeat what you had alread written?

*Note: Duplicate Post deleted - Content Editor*

*PS - Not his fault - hiccup in system when he posted!*

What you can do depends upon your version of UNIX. On some LINUX systems you can use the method described at nice & renice . I would think that for other UNIX implementations one could find appropriate scripts that would do this, at worst a simple script running with admin rights could do this a...

There is no need to quote your original post, it remains visible to all. Also, my complete post should be visible to all users. The important part is does doing a force compile on the original make a difference on the result.

Although my gut tells me that these are interview questions I'll still make an attempt at an answer. the "Kernel" of an OS has a number of settings that can be changed to affect system setup and performance. Examples of kernel settings are such things as how much virtual memory the OS can ...