Fatal Error: Added field has duplicate identifier()

[Tasneem]

Hi all,

I have a complicated job where I look up on 7 tables and the match criteria for all is a Theater_Nr.
The job after being compiled shows this fatal error :Fatal Error: Added field has duplicate identifier(): THEATER_NR [api/schema/schema.C:418]

I have checked for all possibilities I can think of.The only thing I am unsure is of a single LOOKUP STAGE used for all the 7 lookups.Is it because of this I face this error.

Also let me know where can I get the exact message of fatal errors like :[api/schema/schema.C:418]
Thanks in advance.

[ArndW]

Tasneem,

often the cause for this error is using column propagation and re-declaring an incoming coming. Do you have column propagation turned on in any of the output tabs going into this stage?

[vmcburney]

I am going to resurrect this old thread because we are also getting this error and we have opened a case with Ascential support.

There are two Lookup stage errors that abort a job even with a perfectly valid lookup stage. Both errors happen if you have column propogation turned on and make changes to a lookup stage that has already been compiled and run.

The first is an error when you have a column name that is used by both the input link and a reference link.

main program: Fatal Error: Added field has duplicate identifier(): SOURCE_SYSTEM_TEXT_ID [api/schema/schema C:419]

The second error happens if you try to rename a reference column (usually to try and workaround the first error):

LU_CheckElements: Error when checking operator: Key field "SOURCE SYSTEM_TEXT_ID" not found in input schema [keylookup/keylookup C:177]

In both cases the job compiles but aborts when it runs. The easiest fix I have come up with so far (apart from turning off column propogation at the project level) is to delete the lookup and all links in and out of the lookup and add it again. If you get the lookup right the second time around and don't need to modify it then things run fine.

Copying the job or job stages to another job does not fix the problem, it just copies the corrupt links to the new job.