DSXchange

What is this error trying to tell me? What might cause this error?

An index is out of bounds.


Thanks for our help!!!

Hi RStone,

Where is this error coming from?

Giving additional info will help like specific Datastage version, flavor of unix and what stages you are using in your job.

Is it coming from a DB then what is your DB.

I have worked in VB before and "Item is out of bounds" is a very familiar error when trying to read an array element out of the array range.

Give more info so that we can track it down.

--Rich

Thanks for reading. Here is more information

Datastage PX 7.5.1
Unix 5.3

The error happens just after a funnel stage. The funnel stage is reading data from 5 files. One of the files are processed through a trasformer first before going into the funnel. DB2 tables are in the job, but the job aborts before it reaches the DB2 UDB stages.

Can you post the exact error message in its entirety, so we can see the stage type that is generating it? Have you involved your support provider (since you may have uncovered a bug)? Are you using any routines?

Here is the exact error message.

cs}}},1: Fatal Error: An index is out of bounds.

In its entirety please. We need to see what's inside those curly braces. Obfuscate your data if you need to, but not anything produced by the framework. For example, I suspect that the "cs" is a "case sensitive" sub-option to a sort, but can not be sure without the remainder of the information.

That is the entire error message, nothing has been removed or changed. I copied that straight out of the log and pasted it. The error message does not provide much detail.

cs}}},1: Fatal Error: An index is out of bounds.

Yuk. It is clearly an incomplete error message. No information is available about whence it was generated, which makes diagnosis difficult. Are you using non-default partitioning? Is there anything in the Transformer code (in RT_SCnnn directory, where nnn is the job number) that suggests that an array is being (mis-)used?

It is using all default partitioning. It seems the issue is related to a join in the job. If I change the join to a lookup I no longer get this message.