DSXchange

Deleting the record from UV.ACCOUNT is only half the story. You also need to remove the entries from the SQL Catalog. The command for doing this properly is VERIFY.SQL SCHEMA projectname This will report that the physical objects do not exist. You must then log in as a user who has DBA privilege (us...

The "DSU." says that there is an Nvl() function in your Repository. Check which Routines ARE in your job design, and whether any of these might invoke the Nvl() function. Don't forget to check in any before/after subroutines and/or job control routines also.

Another way to look at it is from the database server's perspective.

A running DataStage job, or a metadata import or View Data request from DataStage, is just another client application as far as the database server is concerned.

The "Clear" utility in Director is a DELETE transaction which does not delete the control records, followed by some UPDATE transactions that correct values in those control records. Director must wait until those transactions are completed.

The code is already provided, in a routine called DSSendMailTester (under Routines\Built-In\Utilities in your Repository).

In your ps command it was clear that dsrpcd was not running. You needed to determine why, either by starting it in debug mode, or checking with netstat for processes bound to its port. Search the forum for details. Reinstallation was probably unnecessary.

From a Job Control routine (in any version from 1.0 onwards) Shell = "NT" Command = "Deltxtfiles.bat" ; * or pathname of BAT file Output = "" ExitStatus = -999 Call DSExecute(Shell, Command, Output, ExitStatus) Message = "Executed command: " : Quote...

In the meantime, a manual purge (up to the last run) would help you to get up and running again.

Welcome aboard. :D

Is it the jobs that stop, or the DataStage services? ("DS stop with a strange error" does not let us know which of these things is happening.) Can you please post the full message from that error? Double click on the event in the job log to view the full message.

DS_JOBOBJECTS/DATA.30 is espected to be large, as it contains every object in every job and job sequence design in the project. If can grow towards 2GB before you should become alarmed. As it approaches that size (say, if it exceeds 1.5GB) then you should schedule some down time for maintenance. You...

I must hasten to add that it is not my routine itself that is giving the error! It is the security setting on the routine in its table.

Have you provided a reproducible example to your support provider? If so, what feedback did you get?

Find out who's deleting them. Castigate them.

Run the find command suggested earlier. Something is growing rapidly. Note that auto-purge does not occur for jobs that abort. Auto-purge is triggered only by successful completion (status DSJS.RUNOK or DSJS.RUNWARN).