With our latest release we’ve improved the scheduling functionality in Turbonomic making it much easier to set a schedule for suspending and starting VMs in AWS and Azure.
Below is an example of how to enable this capability with Turbonomic version 6.4 or later.
You will need to:
- Create a Virtual Machine Group for the VMs you would like to suspend and start
- Create an Application Group that correspond to the VMs
- Create the schedules you would like to use. Note, the schedules should only be attached to the Application group
- Set the Automation Policies for the groups
Virtual Machine Group
Create a group of virtual machines under Settings > Groups. In this example I used the AutoSuspend tag is True Filter to define the group.
You will need to create a corresponding Application Group that aligns with the VMs. Note, the Application Group could be larger in scope (e.g. Apps_GuestLoad is a pre-define group that includes all Applications in the public cloud). I used the same tag to define the Application group.
In this example I created two weekly schedules. One for the weekend and one for weekdays. You can find the new scheduling workflow under Settings.
Now we can create the Automation Policies. We will need to create two Application policies – one with the Weekend and one with the Weekday schedule – and one Virtual Machine policy.
Create the Application Policies under Settings>Policies>New Automation Policy>Application. And set the Application Priority to Normal. The schedule should only be attached to the Application Policies. When the schedule expires the Application Priority returns to Mission Critical and the corresponding VM will be prescribed a start action.
For this example, I selected the Application Group I created above and attached the weekend schedule to one policy and the weekday schedule to another.
Finally create the Virtual Machine Policy under Settings>Policies>New Automation Policy>Virtual Machines. In this policy specify the action automation mode (e.g. Manual, Automated) and the Minimum Sustained Utilization under Operational Constraints. Minimum Sustained Utilization compares the value you set to a weighted average of the historical and current VCPU utilization on the VM. If it is below the value you set the VM will suspend.
For this example, I used Automated mode and 99 as the Minimum Sustained Utilization as I want all VMs to suspend and start automatically on a schedule.
That’s it, you are all set. Give this a try in your AWS or Azure environment.