Benchmarking using Control-M

Some information about Autosys
Post Reply
Admin007

Benchmarking using Control-M

Post by Admin007 » 25 Jun 2008 7:25

Currently my company has 5 regions, 4 test regions and 1 Production region running Control-M. None of the test regions are equal, resource wise, jobs that run, etc. to Production. For instance, CPU resources are divided where 90% of the CPU goes to Production, 5% goes to our PStage region and 3% goes to QA, the remaining % is divided between the lower 2 test regions. I need to get Benchmarking in place for comparison testing of jobs before they move to Production. However, since none of the test regions are set up equal to Production, data is skewed and a true benchmark of how a job will run (how long it will take to process) in Production is not feasible at this time. Any ideas on how I could go about Benchmarking jobs that run in a test region to show comparative data for what may occur once the job migrates to Production? Anyone else's company experiencing similar needs? We run about 1800 jobs each in QA and PS but 3000 or more in Production.

User avatar
Dilbert
Nouveau
Nouveau
Posts: 185
Joined: 10 Jan 2007 12:00
Location: Zagreb, Croatia

Re: Benchmarking using Control-M

Post by Dilbert » 25 Jun 2008 8:17

Try using simulation/what-if scenario feature in FOrecast.
Open jobs in forecast and run what-if scenarios. Add a jobs to jor Production in FOrecast and looks what's happening.

Admin007

Post by Admin007 » 25 Jun 2008 8:31

We do not have Forecast. Currently we are running version 6.2. I am in the process of reviewing for an upgrade to v6.3 and I have made a recommendation for Forecast Tool. In the meantime, I need to find a way to benchmark w/ what we have.

User avatar
mauriziog
Nouveau
Nouveau
Posts: 807
Joined: 08 Jun 2007 12:00
Location: Varese - Italy
Contact:

Post by mauriziog » 27 Jun 2008 11:46

We do not have Forecast. Currently we are running version 6.2. I am in the process of reviewing for an upgrade to v6.3 and I have made a recommendation for Forecast Tool. In the meantime, I need to find a way to benchmark w/ what we have
You can work on the statistical data.
You must extract the data and sort them in base of the analisis you whant to do, after this you can analise with eletronic sheets like "ms excel" the results.
The base is the statistical data of exec time, cpu, etc.; after you must build the analisis to have the benchmark.

Post Reply