Home » RDBMS Server » Server Administration » Tracing long running jobs
Tracing long running jobs [message #369529] Fri, 13 October 2000 08:46 Go to next message
Gerard Vaz
Messages: 9
Registered: September 1999
Junior Member
We have a daily end of day job that runs overnight. On some occassions it runs in twice/thrice the amount of time that it normally requires.

I am informed of this early next morning by the operator.

What is the best way to identify/trace what caused the job to run longer that the normal time? Is it possible to identify the bottleneck after the time has elapsed?
Re: Tracing long running jobs [message #369601 is a reply to message #369529] Thu, 02 November 2000 07:59 Go to previous message
simham
Messages: 7
Registered: November 2000
Junior Member
I do not know any tool that answers your question.
but you can as your operator to keep monitoring
the system activities. it is possible that someother process might have taken the priority. also check alert and other trace files for warnings and errors. if a normal process takes longer time, you must look at the other activities of the mechine. like load on cpu,load on memory,load on i/o et cetera.
Previous Topic: Oracle8.1.6 doesn't work well
Next Topic: ORA-27123: unable to attach to shared memory segment
Goto Forum:
  


Current Time: Thu May 02 01:17:19 CDT 2024