vCPU Capacity Counting Methodology for IBM Programs
- The vCPU capacity* of a pod* is the sum of the CPU limits* for all containers within that pod. Pods with vCPU capacity less than one core will be counted as fractions.
- Within a worker node*, when the total vCPU capacity for all pods of an IBM Program exceeds the worker node capacity, the vCPU capacity is equal to the worker node capacity.
- vCPU capacity for an IBM Program will be aggregated at the cluster level* and any fractional values rounded up to the nearest whole integer.
- For IBM Cloud Paks and other IBM Programs that are Bundled Solutions of multiple products with license ratios, the counting methodology applies to each Bundled Product
*See definitions below.
- Cluster - A set of worker machines, called nodes, that run containerized applications. Every cluster has at least one worker node.
- Container - A lightweight and portable executable image that contains software and all of its dependencies.
- CPU limit - Provides constraints to limit resource consumption per Containers in a namespace via the use of the Kubernetes CPU limits property “resources.limits.cpu.”
- Namespace - An abstraction used by Kubernetes to support multiple virtual clusters on the same physical cluster
- Pod - The smallest and simplest Kubernetes object. A Pod represents a set of running containers on your cluster.
- VPC (Virtual Processor Core) & PVU (Processor Value Unit) - Units of measure by which Programs can be licensed (see the Common License Metrics page for full definitions). For license counting purposes, IBM counts vCPUs, adjusts based on Simultaneous multithreading (SMT) or Hyperthreading (HT) ratios. (See Additional License Counting Guidance section for more information), and then converts to the license metric.
- vCPU capacity - The total number of vCPUs available to a Pod as defined by the counting methodology (below). Please note: Any pod which contains a container without a CPU limit is considered to have vCPU capacity equal to the worker node capacity
- Worker node - A worker node is a node that provides a containerized environment for running tasks. As demands increase, more worker nodes can be added to your cluster to improve performance and efficiency. A cluster can contain any number of worker nodes, but a minimum of one worker node is required.
- Worker node capacity – The total capacity available to the worker node, measured by either virtual cores or physical cores, as reported by Kubernetes API.
Additional License Counting Guidance
- 2 vCPUs per 1 core as 1 VPC or 70 PVUs for license counting purposes if Hyperthreading or SMT2 is enabled (e.g., 8 vCPUs is counted as 4 vCPUs).
- 4 vCPUs per 1 core as 1 VPC or 70 PVUs for license counting purposes if SMT4 is enabled (e.g., 16 vCPUs is counted as 4 vCPUs).
- 8 vCPUs per 1 core as 1 VPC or 70 PVUs for license counting purposes if SMT8 is enabled (e.g., 32 vCPUs is counted as 4 vCPUs).
- For clusters, with mixed/different SMT levels, the lowest factor is applied. E.g.: some nodes use SMT4 and some other use SMT8, factor of 4 will be used for this cluster's license calculation. Identification of the HT/SMT status is to be done manually and assigned at a cluster level.
As a note: Until an automatic adjustment functionality is integrated into IBM License Service, the adjustment is to be made manually to the IBM License Service output. Once the functionality is available, users are to ensure that they are on the latest version of IBM License Service and their clusters are configured as described in the product documentation.
- For some IBM Programs, configuration options and user activities can increase the license requirement due to additional workload being created.
- Customer must license all workload created by IBM Program, including workload resulting from changes in product configuration and dynamically created workload generated by user activity.
- Customers are responsible for licensing Red Hat OpenShift subscriptions in accordance with Red Hat’s licensing agreement. Please refer to each IBM Cloud Pak’s licensing information document for more information.
- Customers are responsible for controlling license demand through a combination of Red Hat® OpenShift® capabilities such as namespace* quotas and IBM Program specific capabilities such as deployment options and scaling controls. For IBM Program specific information on controlling license demand, refer to the corresponding IBM Program documentation
- The preset resources.limits.cpu settings that ship with each Program are used in all QA, performance testing, etc. While IBM does not recommend modifying these settings, if you should choose to do so, please first consult the product documentation to understand the acceptable documented ranges for your Programs
Note: If you have questions about licensing requirements, please contact your IBM representative for additional support.
IBM License Service
To take advantage of container licensing, IBM License Service must be used to track license usage and determine your required entitlement. There are no exceptions to this rule.
The IBM License Service is the only accepted tool for license usage tracking and reporting container licensing. It is:
- contained in each IBM Cloud Pak’s services
- available for download for use in IBM Certified Containers by contacting firstname.lastname@example.org (or product support). Customers should verify whether IBM License Service is installed and configured correctly when deploying their containerized Programs
Available vCPU capacity will be polled at frequent intervals (≤ 30 minutes) throughout each day and processed where the maximum value from intraday calculations determine daily available vCPU capacity
It is essential that you use and maintain the license usage tracking tool (IBM License Service: see below) included in each of the Cloud Pak services. This service will track license usage and determine your required entitlement needs.
Additional Compliance & Reporting requirements are listed in the transaction addendum stating container licensing terms
Here are just a few learn more resources you may wish to explore
- IBM Container Licensing FAQs
- IBM Learning Hub > Containers
- IBM Cloud Paks
- IBM Containers-as-a-Service
- Container Management 2.0 Next-Gen Platforms Package Management, Orchestration, and Services for a Streamlined User Experience (Frost & Sullivan Stratecast)
- Containers: Automate the deployment, scaling, and management of containerized applications