Kubernetes has become a standard deployment model for cloud-native computing workloads, but figuring out how to manage and optimize costs hasn't always been easy.
On June 2, Kubecost announced the launch of the open source OpenCost project — a multi-stakeholder effort to provide Kubernetes cost monitoring and optimization capabilities.
Kubernetes is a technology for container and microservices orchestration, operating as an open source project governed by the Cloud Native Computing Foundation (CNCF). The new OpenCost project is being submitted to the CNCF for consideration as a new project.
OpenCost benefits from the support of multiple vendors, including Adobe, Armory, AWS, D2iQ, Google, Mindcurv, New Relic, and SUSE, alongside Kubecost.
"At Kubecost, we see firsthand that many different platforms and Kubernetes providers have different ways of tracking costs — including different methodologies for allocations, costs, and terminology," Webb Brown, co-founder and CEO of Kubecost, told ITPro Today. "We hear this consistently from our end users as a point of friction for them."
How OpenCost Aims to Improve Kubernetes Cost Monitoring
For teams running across environments, the lack of consistency can be a source of friction for Kubernetes adoption, according to Brown.
The lack of standardization makes life more difficult for vendors because users are constantly requesting different cost tracking methodologies from them, Brown said. As such, there was a natural opportunity for Kubecost to collaborate with both end-user organizations and cloud providers on the OpenCost project, he said.
"So many large organizations are now using Kubernetes, and solving the complexities around measuring and allocating costs is becoming a business-critical challenge for them," Brown said. "As we enter a period of more unknown economic circumstances, I expect cost tracking to become even more important for a wide range of organizations."
How OpenCost Supports Kubernetes Cost Management
OpenCost combines the cost allocation engine of Kubecost with a brand-new specification, built in collaboration with the founding partners of the OpenCost project, according to Brown. Both the cost allocation engine and the specification have been submitted to the CNCF.
OpenCost ingests data from cloud APIs, on-premises environments, and the Kubernetes scheduling APIs.
"The goal of OpenCost is to be vendor-agnostic, and that means we need to ingest data and cost insights from as many platforms and providers as possible," Brown said. "Today, OpenCost can run anywhere Kubernetes is running, and we will continue to build on OpenCost while keeping vendor neutrality as a core tenet."
Among the capabilities that Brown hopes to be built in OpenCost in the coming months and years is support for more cloud providers as well as more features, including exposing the carbon emission ratings and environmental cost of clusters and workloads.
"We are really excited about opening OpenCost to the broader community and getting as many contributions from different organizations and end users as possible," he said. "The more diverse the contributions are, the more we can make OpenCost useful to a very broad set of end users."