GKE overview

Autopilot Standard

This page describes Google Kubernetes Engine (GKE), a managed Kubernetes service that you can use to deploy and operate containerized applications at scale using Google's infrastructure. This page is intended for platform administrators who are looking for a scalable, automated, managed Kubernetes solution. You should already be familiar with Kubernetes concepts.

GKE is a Google-managed implementation of the Kubernetes open source container orchestration platform. Kubernetes was developed by Google, drawing on years of experience operating production workloads at scale on Borg, our in-house cluster management system.

When to use GKE

GKE is ideal if you need a platform that lets you configure the infrastructure that runs your containerized apps, such as networking, scaling, hardware, and security. GKE provides the operational power of Kubernetes while managing many of the underlying components, such as the control plane and nodes, for you.

Benefits of GKE

The following table describes some of the benefits of using GKE as your managed Kubernetes platform:

GKE benefits
Platform management	Fully-managed nodes in GKE Autopilot mode with built-in hardening and best practice configurations automatically applied. Managed upgrade experience with release channels to improve security, reliability, and compliance. Flexible maintenance windows and exclusions that let you configure upgrade type and scope to meet business needs and architecture constraints. In GKE Standard mode, flexible node upgrade strategies to optimize availability and manage disruptions. Automatic scaling of nodes based on the number of Pods in the cluster with Autopilot mode or with node auto-provisioning in Standard mode. Node auto-repair to maintain node health and availability. Built-in logging and monitoring. Google Cloud integrated CI/CD options with Cloud Build and Google Cloud Deploy.
Improved security posture	Hardened node operating system for apps: Container-Optimized OS. Built-in security measures. Automatic upgrades to new GKE versions. Integrated security posture monitoring tooling with the security posture dashboard. Google Cloud logging and monitoring integrations with Google Cloud's operations suite.
Cost optimization	In Autopilot mode, pay only for the compute resources your running Pods request. In GKE Standard mode, you pay for all resources on nodes, regardless of Pod requests. Save costs by running fault-tolerant workloads, such as batch jobs, on Spot Pods. Minimized operational overhead in Autopilot mode because Google manages both the nodes and the control plane.
Reliability and availability	>99% monthly uptime SLO. Pod-level SLA in Autopilot clusters because Google manages the nodes. Highly-available control plane and worker nodes in Autopilot mode and in regional Standard clusters. Proactive monitoring and recommendations to mitigate potential workload disruptions caused by upcoming deprecations. Multi-cluster Service capabilities.

Use cases for GKE

GKE and Kubernetes are used in a variety of industries, including robotics, healthcare, retail, education, gaming, and financial services. The range of applications include AI and ML operations, data processing at scale, operating scalable online games platforms, and running reliable applications under heavy load.

For case studies by industry and application, refer to Google Cloud customers.

How GKE works

A GKE environment consists of nodes, which are Compute Engine virtual machines (VMs), that are grouped together to form a cluster. You package your apps (also called workloads) into containers. You deploy sets of containers as Pods to your nodes. You use the Kubernetes API to interact with your workloads, including administering, scaling, and monitoring.

Kubernetes clusters have a set of management nodes called the control plane, which run system components such as the Kubernetes API server. In GKE, Google manages the control plane and system components for you. In Autopilot mode, which is the recommended way to run GKE, Google also manages your worker nodes. Google automatically upgrades component versions for improved stability and security, ensuring high availability, and ensuring integrity of data stored in the cluster's persistent storage.

For more information, refer to GKE cluster architecture.

Kubernetes versions and features

GKE automatically upgrades your control plane to new Kubernetes versions that add new features and improvements in the open source project. The Kubernetes version selected for auto-upgrades depends on the stable version in the GKE release channel you select when you create the cluster. You can also manually upgrade your control plane to a different Kubernetes version than the version GKE selects for an upgrade. For detailed information on versions and upgrades, refer to the release notes and GKE versioning and upgrades. If you use GKE Standard mode and don't enroll in a release channels, you won't get automatic upgrades.

GKE includes most beta and stable Kubernetes features. If you want to try less stable Kubernetes features in the alpha stage, use alpha Standard clusters.

Modes of operation

GKE has the Autopilot and Standard modes of operation, which offer you different levels of flexibility, responsibility, and control. We recommend the fully-managed Autopilot mode, in which Google Cloud manages your nodes for you and provides a workload-focused, cost-optimized, production-ready experience.

If you want more information before you choose a mode, refer to Choose a GKE mode of operation.

Get started with GKE

To start using GKE, try the quickstart to deploy a containerized web application. Then, read the Autopilot overview, which has guidance and resources for planning and operating your platform.

What's next

Choose a GKE mode of operation