Skip to content

Latest commit

 

History

History
157 lines (115 loc) · 6.56 KB

README_en.md

File metadata and controls

157 lines (115 loc) · 6.56 KB
Raw

Awesome Build Status

Introduction

Build a production kubernetes HA cluster.

  • Every node config a ipvs proxy for masters LB, so we needn't haproxy or keepalived any more.
  • Then run a lvscare as a staic pod to check apiserver is aviliable. /etc/kubernetes/manifests/sealyun-lvscare.yaml
  • If any master is down, lvscare will remove the ipvs realserver, when master recover it will add it back.
  • Sealos will send package and apply install commands, so we needn't ansible.

✨ Supported Environment

Linux Distributions, CPU Architecture

  • Debian 9+, x86_64/ arm64
  • Ubuntu 16.04， 18.04， 20.04 , x86_64/ arm64
  • Centos/RHEL 7.6+, x86_64/ arm64
  • 99% systemd manage linux system， x86_64/ arm64
  • Kylin arm64

kubernetes Versions

  • 1.16+
  • 1.17+
  • 1.18+
  • 1.19+
  • 1.20+
  • 1.21+
  • 1.22+

Looking for more supported versions, sealyun.com. sealos is currently supported the latest k8s 1.22+

Requirements and Recommendations

  • Minimum resource requirements

    • 2 vCpu
    • 4G RAM
    • 40G+ Storage

  • OS requirements

    • SSH can access to all nodes.
    • hostname is unique, and satisfied kubernetes requirements.
    • Time synchronization for all nodes.
    • if network card has a stranger name, change it to (eth.|en.|em.*).
    • kubernetes1.20+, use containerd for default cri. user should not to install containerd or docker-ce. sealos will do it.
    • kubernetes1.19-, use docker for default cri. user should not to install docker-ce. sealos will do it for you.

  • Networking and DNS requirements：

    • Make sure the DNS address in /etc/resolv.conf is available. Otherwise, it may cause some issues of DNS in cluster.
    • if you use aliyun/huawei cloud to deploy kubernetes. default pod cidr is conflict with dns cidr, we recommend you install kubernetes init flag to add --podcidr to avoid this problem.
    • sealos default to disable firewalld, It's recommended that you turn off the firewall. if you want to use firewalld , remember to allow kubernetes port traffic.

  • Kernel requirements:

    • When cni components choose cilium, the kernel version must not be lower than 5.4

🚀 Quick Start

Environmental information

Hostname IP Address
master0 192.168.0.2
master1 192.168.0.3
master2 192.168.0.4
node0 192.168.0.5

Server password：123456

kubernetes .0, the version is not recommended for production environment!!!

Just prepare the server and execute the following command on any server

# download and install sealos, sealos is a binary tool of golang, just download and copy directly to the bin directory, the release page can also be downloaded
$ wget -c https://sealyun.oss-cn-beijing.aliyuncs.com/latest/sealos && \
    chmod +x sealos && mv sealos /usr/bin

# download offline resource pack
$ wget -c https://sealyun.oss-cn-beijing.aliyuncs.com/05a3db657821277f5f3b92d834bbaf98-v1.22.0/kube1.22.0.tar.gz

# Install a three-master kubernetes cluster
$ sealos init --passwd '123456' \
	--master 192.168.0.2  --master 192.168.0.3  --master 192.168.0.4  \
	--node 192.168.0.5 \
	--pkg-url /root/kube1.22.0.tar.gz \
	--version v1.22.0

Parameter meaning

parameter meaning example
passwd server password 123456
master k8s master IP Address 192.168.0.2
node k8s node IP Address 192.168.0.3
pkg-url offline resource package address, support downloading to local or a remote address /root/kube1.22.0.tar.gz
version Resource pack Corresponding version v1.22.0

add master

🐳 → sealos join --master 192.168.0.6 --master 192.168.0.7
🐳 → sealos join --master 192.168.0.6-192.168.0.9  # or multiple consecutive IPs

add node

🐳 → sealos join --node 192.168.0.6 --node 192.168.0.7
🐳 → sealos join --node 192.168.0.6-192.168.0.9  # or multiple consecutive IPs

delete the specified master

🐳 → sealos clean --master 192.168.0.6 --master 192.168.0.7
🐳 → sealos clean --master 192.168.0.6-192.168.0.9  # or multiple consecutive IPs

Delete the specified node

🐳 → sealos clean --node 192.168.0.6 --node 192.168.0.7
🐳 → sealos clean --node 192.168.0.6-192.168.0.9  # or multiple consecutive IPs

clean up the cluster

🐳 → sealos clean --all

✅ Feature

  • Support ARM version offline package, v1.20 version offline package supports containerd integration, completely abandon docker
  • 99 years certificate, support cluster backup and upgrade
  • Does not rely on ansible haproxy keepalived, a binary tool, 0 dependencies
  • Offline installation, different versions of kubernetes download corresponding to different versions Resource pack, Offline package contains all binary files configuration files and images
  • High-availability local LIB implemented through ipvs, which takes up less resources, is stable and reliable, and is similar to the implementation of kube-proxy
  • Almost compatible with all environments that support systemd x86_64 architecture
  • Easily add and delete cluster nodes
  • Thousands of users use sealos in the online environment, which is stable and reliable
  • The resource pack is placed on Alibaba Cloud OSS, so you don’t have to worry about network speed anymore
  • dashboard ingress prometheus apps offline packaging, a key installation

📊 Stats

Alt

简体中文

More offline packages