Workload Policies

Control-M’s Best Kept Secret

In my travels around the country visiting with Control-M customers, I find so many that are not aware of the one addition to the product that is beneficial to an operations/NOC team. In version 7 BMC introduced Workload Policies, designed for controlling workload when the active batch is having issues.

USING QUANTITATIVE RESOURCES

Most customers would use Quantitative Resources to control workload. This option has been around for years and is ingrained into the scheduling culture. For those who do not understand Quantitative Resources, they are assigned to each job you want to have control over. A Quantitative Resource called SAP could be defined to all SAP jobs. It is used to control how many jobs can run simultaneously. Sometimes you want all SAP jobs to run, others you want to control the number or set it to 0 to stop jobs from running. Quantitative Resources work fine if you know in advanced what is going to happen.

WORKLOAD POLICIES – AN ALTERNATIVE TO QUANTITATIVE RESOUCE

But what if you do not have Quantitative Resources on every job? How do I stop all jobs from taking off, what about making sure that SAP jobs don’t run during maintenance? How do I move jobs running on AGENTA to AGENTB? How can I quickly control the max available in my Quantitative Resource?

All of these and more can be controlled by Workload Policies. Policies are defined, then can be activated or deactivated with ease. There is a command line to do this as well.

The most common uses for Workload Policies are to dynamically stop work from running and to control Quantitative Resources. No Quantitative Resources are required for Workload Policies to take effect. When you active the policy below it will stop all Informatica jobs from being submitted. If you had multiple Informatica CM’s you could stop them for a specific agent.

Notice that while we used any Informatica job running on a specific Control-M Server, you could have used the other fields as well. So maybe use jobname starting with prod_hb1 or application HB, any of the fields could be used. The job type has a drop down field with all of your job types listed, including any that were created using Application Integrator.

SCHEDULING OF WORKLOAD POLICY

Another feature that can dramatically reduce the time spent trying to match specific scenarios is scheduling the policy to automatically start and stop. You can select specific dates or days like Monday, Wednesday, and Saturday. You can specify date ranges as well.

For the time, you can select a specific time or use the between times to prevent jobs from running during a maintenance window.

HOST MAPPING POLICY

The last topic to cover for Workload Policies is the

ability to have jobs with a specific agent in the Hostname field be able move to a different agent or agentless server.

In the following example, we set up a policy for all jobs running on ServerA. The scenario would be that ServerA is having some kind of issues and they want you to move all of the jobs to ServerB. The policy will cover that move for as long as it is active.

You can create these policies and leave them out there waiting to be activated. To activate click on the “Activate” icon on the top of the GUI or do a right click and select Activate.

JOBS IN AJD REQUIRING POLICY

Just like when jobs are waiting for resources of any kind, the color status of a job waiting on a workload policy is blue. Looking at the Wait tab shows that the job is waiting for a workload policy.

SUMMARY

Hopefully this has helped show the benefits of using Workload Policies in Control-M. It offers significant options to controlling active workload. It dramatically decreases the time spent trying to hold or redirect work as well.

Again, this functionality has been around since V7 and is an extremely helpful tool for your Operations/Noc teams.

Visit the BMC Control-M Community page for more information on this and other topics.