Kruise

English | 简体中文

What is NEW!
Sep 6th, 2021. Kruise v0.10.0 is RELEASED! It provides new controllers like WorkloadSpread and PodUnavailableBudget, please check the CHANGELOG for details.
May 20th, 2021. Kruise v0.9.0 is RELEASED! It provides great features such as ContainerRecreate and DeletionProtection, please check the CHANGELOG for details.
Mar 4th, 2021. Kruise v0.8.0 is RELEASED! It provides refactoring SidecarSet, Deployment hosted by UnitedDeployment, and a new kruise-daemon component which supports image pre-download, please check the CHANGELOG for details.

Introduction

OpenKruise (official site: https://openkruise.io) is now hosted by the Cloud Native Computing Foundation (CNCF) as a Sandbox Level Project. It consists of several controllers which extend and complement the Kubernetes core controllers for workload and application management.

Key Features

Typical Workloads

Typical Workloads can help you manage applications of stateless, stateful and daemon.

They all support not only the basic features which are similar to the original Workloads in Kubernetes, but also more advanced abilities like in-place update, configurable scale/upgrade strategies, parallel operations.
Job Workloads
- BroadcastJob for deploying jobs over specific nodes
- AdvancedCronJob for creating Job or BroadcastJob periodically
Sidecar container Management

The Sidecar containers can be simply defined in the SidecarSet custom resource and Kruise will inject them into all Pods matched.

The implementation is done by using Kubernetes mutating webhooks, similar to what istio does. However, it allows you to explicitly manage your own sidecars.
- SidecarSet for defining and upgrading your own sidecars
Multi-domain Management

This can help you manage applications over nodes with multiple domains, such as different node pools, available zones, architectures(x86 & arm) or node types(kubelet & virtual kubelet).

Here we provide two different ways:
- WorkloadSpread for bypass distributing pods in workloads
- UnitedDeployment, a new workload to manage multiple sub-workloads
Enhanced Operations
- Restart containers in a running pod
- Download images on specific nodes
Application Protection
- Protect Kubernetes resources and applications' availability from the cascading deletion
- PodUnavailableBudget for achieving the effect of preventing application disruption or SLA degradation

Quick Start

We strongly recommend you to use Kruise with Kubernetes version >= 1.16. For these clusters, you can simply install Kruise with helm v3.1.0+:

helm install kruise https://github.com/openkruise/kruise/releases/download/v0.10.0/kruise-chart.tgz

Note that installing this chart directly means it will use the default template values for the kruise-manager. You may have to set your specific configurations when it is deployed into a production cluster or you want to configure feature-gates.

For more install/upgrade details or older Kubernetes versions, please read this doc.

Documentation

You can view the full documentation from the OpenKruise website.

Users

Registration: Who is using Kruise

Alibaba Group, Ant Group, DouyuTV, Sto, Boss直聘
hangyinxiaofei, vanyitech, Dmall, Bringg, 佐疆科技
Lyft, Ctrip, 享住智慧, VIPKID, zhangmen
xiaohongshu, bixin, 永辉科技中心, 跟谁学, 哈啰出行
Spectro Cloud, ihomefnt, Arkane Systems, Deepexi, 火花思维
OPPO, Suning.cn, joyy, Mobvista

Contributing

You are warmly welcome to hack on Kruise. We have prepared a detailed guide CONTRIBUTING.md.

Community

Active communication channels:

Slack: OpenKruise channel (English)
DingTalk：Search Group ID 23330762 (Chinese)
Bi-weekly Community Meeting (APAC, Chinese):
- Thursday 19:00 GMT+8 (Asia/Shanghai), Calendar
- Meeting Link(zoom)
- Notes and agenda
Bi-weekly Community Meeting (English): TODO

License

Kruise is licensed under the Apache License, Version 2.0. See LICENSE for the full license text.

yan234280533/kruise