Introduction to CPU Speeds (vCPU Classes)

Content / Solution:

Overview of CPU Speeds (MCP 2.0 Locations Only)

If the infrastructure in a given User-Manageable Cluster or MCP 2.0 Data Center offers the ability to deploy vCPU with different performance characteristics, the CloudControl software will allow you to decide which vCPU class a given Cloud Server should utilize. CPU Speeds allow the user to select an appropriate level of performance based on the intended function of the Cloud Server. For example, Cloud Servers may be running applications that are very CPU-dependent and need more consistent CPU performance. The performance characteristic associated with the CPU "Speed" applies to all vCPU on a given Cloud Server.

CPU Speeds that may be available in a given User-Manageable Cluster or Data Center include the following:

• Standard - Standard is the "default" level of CPU Speed and matches the level of performance available prior to the introduction of CPU Speeds. It is available in all Clusters and Data Centers. All Cloud Servers introduced prior to the introduction of the CPU Speed feature have been assigned "Standard" CPU Speed. In Public CaaS environments, each vCPU running on "Standard" CPU Speed can burst up to a maximum of 1.5 GHz, depending on availability on the specific processor at the time of request. All MCP 1.0 locations only offer a single "Standard" CPU Speed.
• High Performance - The "High Performance" CPU Speed provides both a more consistent vCPU performance and a higher "burstable" threshold. This CPU Speed is available in all Public MCP 2.0 locations. In Public CaaS, each vCPU has three times the hypervisor "reservation" (for more consistent performance) and can burst up to 2.187 GHz, depending on availability on the specific processor at the time of request.

Each CPU Speed is tracked separately for usage purposes and carries different hourly pricing. Usage for the "Standard" CPU Speed is tracked under the "CPU Hours" pricing element. The "High Performance" CPU Speed is tracked based on "High Performance CPU Hours" element.

The  CPU Speed associated with each Server is visible in the Server information area:

The speed can be managed via the Manage Server dialog of the Admin UI in the Server Specs tab as described in How to Manage a Cloud Server 

Availability and Use of Different CPU Speeds

Not all User-Manageable Clusters and MCP 2.0 Data Center Locations have the same CPU Speeds available. Only the "Standard" CPU Speed is available in all Clusters and Data Centers. To identify what CPU Speeds are available in a given Cluster or Data Center, view the summary information as described in How do I Identify Hardware Specifications and Capabilities Available in a Data Center Location

It is possible, although unusual, for a CPU Speed to exist, but not currently be available for use. If such a condition exists, the fact that the speed isn't currently available and the reason will be displayed in the same dialog. 

How to Use CPU Speeds

There are two ways to set the CPU Speed for a Cloud Server:

• You can choose the CPU Speed when a Server is deployed. By default, the Admin UI will use the CPU Speed associated with a Client Image. However, you can modify that CPU Speed through a dialog on the Server as described in How to Deploy a Cloud Server from a Guest OS Customization Image  
• Client Images inherit the CPU Speed of the Cloud Server when the Image was created. Imported Images are always imported with the "Standard" CPU Speed. You can edit the CPU Speed associated with a Client Image as described in How to Manage Client and Long-Term Retention Images
  • Note: Client Images lose any CPU Speed characteristics when a Client Image is exported since the OVF format does not support the concept. This means that all Imported Images will have all CPU Speeds set to Standard.
• Once a Server is deployed, you can change the CPU Speed as described in How to Manage a Cloud Server 

As always, these functions are also available programmatically as described in the API documentation at API 2

Reporting and Billing for Different CPU Speeds

Each unique CPU Speed will be charged at its own rate for the pricing element. Refer to your Cloud product pricing page for more information.

From a reporting perspective, the system will now provide separate reporting for each CPU Speed that's available within a Geographic Region. This impacts the reports as follows:

• The Summary Usage Report:
  • The CPU Hours column tracks the usage of Standard CPU Speed.
  • An additional column is added for each additional CPU Speed available in a given Geographic Region.
  • See How to Create a Summary Usage Report for details.
• The Detailed Usage Report.
  • The CPU Hours column tracks the usage of Standard CPU Speed.
  • An additional column is added for each additional CPU Speed available in a given Geographic Region.
  • See How to Create a Detailed Usage Report
• Columns will be present for each CPU Speed available in any data center location in a Geographic Region for all accounts in the Geographic Region. This reporting will appear regardless of whether any of the account's servers are actually using these CPU Speeds or whether the account even has access to a data center location supporting the CPU Speed.