Cutting cluster management fees on Google Kubernetes Engine
Director of Product Management
Today, we're excited to announce that we have eliminated the cluster management fee for Google Kubernetes Engine, our managed Kubernetes service.
We founded the Kubernetes open-source project in 2014, and have remained the leading contributor to it. Internally at Google, we’ve been running globally scaled, production workloads in containers for over a decade. Kubernetes and Kubernetes Engine include the best of what we have learned, including the advanced cluster management features that web-scale production applications require. Today’s announcement makes Kubernetes Engine’s cluster management available at no charge, for any size cluster, effective immediately.
To put this pricing update in context, Kubernetes Engine has always provided a managed master at no charge for clusters of fewer than six nodes. For larger clusters we also provided the managed master at no charge, however we charged a flat fee of $0.15 per hour to manage the cluster. This flat fee is now eliminated for all cluster sizes. At Google, we’ve found that larger clusters are more efficient — especially when running multiple workloads. So if you were hesitating to create larger clusters worry no more and scale freely!
|Nodes in the cluster||Older Pricing||New Pricing (effective immediately)|
|Cluster Management Fee||Cluster Management Fee|
|0 to 5 nodes||0||0|
|6+ nodes||$0.15 / hour||0|
That’s great news, but some of you may be wondering what all is included in cluster management? In the context of Google Kubernetes Engine, every cluster includes a master VM that acts as its control plane. Kubernetes Engine’s cluster management includes the following capabilities among others:
- Master VM is automatically scaled, upgraded, backed up and secured
- Master VM is fully managed by Google Site Reliability Engineering (SRE)
- Advanced features that let you choose to automatically repair, upgrade and scale your entire cluster
- Kubernetes Engine comes with a Service Level Agreement (SLA)
A useful point of comparison is the cost of managing your Kubernetes cluster yourself, either on Google Compute Engine or on another cloud. In a self-managed cluster, you pay for the VM that hosts the master and any resources you need for monitoring, logging and storing its state. Depending on the size of your cluster, moving to Kubernetes Engine could save a decent fraction of your total bill just by saving the cost of the master.
Of course while dollar savings are nice, we have invested Google engineering in automating cluster management with Kubernetes Engine to you save time and headaches as well. In a self-managed cluster, you're responsible for scaling the master as your cluster grows, and for backing up etcd. You have to keep an eye out for security patches and apply them. To access new Kubernetes features, you have to upgrade the master and cluster yourself. And most likely cluster repair and scaling is manual. With Google Kubernetes Engine, on the other hand, we take care of all of this complexity at no charge so you can focus on your business.
[Google Kubernetes Engine] gives us elasticity and scalable performance for our Kubernetes clusters. It’s fully supported and managed by Google, which makes it more attractive to us than elastic container services from other cloud providers
— Arya Asemanfar, Engineering Manager at Mixpanel
We’re committed to raising the bar on Kubernetes’ reliability, cost-effectiveness, ease-of-use and enterprise readiness, and continue to add advanced management capabilities into Kubernetes Engine. For a preview of what’s next we invite you to join an early access program for node auto-provisioning, a new cluster management feature that provisions the right type of nodes in your auto-scaling cluster based on the observed behavior of your workloads. To join the early access program, fill out this form.