Instance with vCPU Pinning

When running latency-sensitive applications inside a virtual machine, you can use virtual CPU pinning to eliminate the extra latency that virtualization imposes.

Important: This feature is offered in VMware Integrated OpenStack Carrier Edition only. For more information, see VMware Integrated OpenStack Licensing.

Virtual CPU pinning enables high latency sensitivity and ensures that all memory and an entire physical core are reserved for the virtual CPU of an OpenStack instance. You configure virtual CPU pinning on a flavor and then create instances with that flavor.

To enable high latency sensitivity for selected virtual CPUs within a VM, you can use flavor extra specs. See Supported Extra Specs for Flavor.

Procedure

1. Log in to the VMware Integrated OpenStack dashboard as a cloud administrator.
2. Select the admin project from the drop-down menu in the title bar.
3. Select Admin > Compute > Flavors
4. Create a flavor or choose an existing flavor to use for virtual CPU pinning.
5. Select Update Metadata next to the flavor that you want to use.
6. In the Available Metadata pane, select and configure the required metadata.
   1. Expand CPU Pinning and click the Add (plus sign) icon next to CPU Pinning policy.
   2. Set the value of hw:cpu_policy to dedicated.
   3. Expand VMware Policies and click the Add (plus sign) icon next to VM latency sensitivity.
   4. Set the value of vmware:latency_sensitivity_level to high.
   5. Expand VMware Quota and click the Add (plus sign) icon next to CPU Reservation in Percentage and Memory Reservation in Percentage.
   6. Set the value of quota:cpu_reservation_percent and quota:memory_reservation_percent to 100.
7. Click Save.

What to do next

You can now enable virtual CPU pinning on an instance by configuring it with the flavor that you modified in this procedure.