IBM Looks to Make Container Pricing Consistent

IBM has moved to standardize the way it charges licensing fees for container technologies based on the number of virtual CPUs consumed.

IT organizations for all IBM Certified Containers must acquire entitlements equal to the license requirements determined by the vCPU Capacity Counting Methodology, as measured by the IBM License Service that IBM employs to track usage.

Similarly, IT teams must also acquire entitlements equal to the license requirements as derived in the vCPU Capacity Counting Methodology and measure by the IBM License Service for IBM Cloud Paks, which are pre-configured stacks of middleware software that IBM makes available on the Red Hat OpenShift platform. Each Cloud Pak has its own unique conversion ratio. In the absence of a ration, IT teams can assume a 1:1 ratio between the Cloud Pak and vCPU capacity.

Charles Quincy, program director for portfolio management at IBM, says IBM is attempting to bring some clarity to the cost of implementing containers. There’s a lot of interest in the flexibility and agility that containers enable. However, many IT teams need know exactly what their IT costs will be, he says.

Developers don’t always carefully consider costs when spinning up IT infrastructure. It will be up to each IT team to decide how many microservices based on containers per virtual CPU, he says. The more microservices per virtual CPU, the lower costs are. However, many organizations quickly encounter performance issues when too many microservices share the same virtual CPU instance. Before too long, they need to spin up additional virtual CPUs to maintain application service level agreements (SLAs).

In the wake of the economic downturn brought on by the COVID-19 pandemic, there’s a lot more focus on cloud costs, especially when it comes to deploying workloads in the cloud. However, the containers that cloud-native applications are based on are often ephemeral. That makes it challenging to predict costs. In an ideal world, IT leaders naturally would prefer to reduce costs.

In the absence of being able to achieve that goal, it’s now absolutely essential that at the very least cost predictions are accurate. That can be tough to gauge when workloads are spinning up and down in ways that are difficult to predict. IBM is betting a pricing model based on a traditional virtual CPU model will at the very least provide more certainty, notes Quincy.

It’s not clear to what degree uncertainty over costs may be hampering the transition to cloud-native platforms. IT organizations that are resistant to change will always bring up cost concerns as a reason to not embrace any emerging technology.
Eventually, however, every IT platform is held to account. It’s not clear if there is an alternative to software licenses based on virtual CPUs that is more viable. Chances are good, however, there will be a lot of experimentation ahead that may prove to be less or more expensive depending on the unique use case of each organization.

Mike Vizard

Mike Vizard is a seasoned IT journalist with over 25 years of experience. He also contributed to IT Business Edge, Channel Insider, Baseline and a variety of other IT titles. Previously, Vizard was the editorial director for Ziff-Davis Enterprise as well as Editor-in-Chief for CRN and InfoWorld.

Mike Vizard has 1606 posts and counting. See all posts by Mike Vizard