DSXchange

Hello,

i need some help in finding the reason for an "abnormal termination" in one job.
The jobs uses ODBC stages and the way i found to resolve this problem was to substitute ODBC stages by OCI stages.
The original job uses ODBCs as input to a Transformer that uses ODBCs and Hashfile as lookups. The output are .dat files and ORABULKs with 'After-stage subroutine'.
The core file generated has this information: "core: ELF-64 core file from 'uvsh' - received SIGINT" that indicates the problem is in the job.
The version of DataStage is 7.5.1A and the server is an HP-UX 11iv2 64 bit.
Has anyone ever get this message (Abnormal termination of stage...)?

Thank you

Pedro

When we were using the HP-UX as a ASCL server (7.5) we used to get this error when the server was full of connection. I think we were able to just connect 5 connections (clients) and more than 5 it would give us this error as well sometimes threw us out of connection.

May be you will like to increase the tracing level and see whther it is this problem or any other problem than this.

That the underlying engine (uvsh) has generated a SIGINT doesn't help you to diagnose the problems; it uses signals for very many things. This requires specialist analysis; preserve the core file and report the problem to your support provider. They may need to analyse the core file.

You may find a CORE command that you can execute from the Administrator client Command window or from a telnet session (not on Windows operating systems). This command does not analyze core files; it simply reports statistics about your current memory usage.