简体中文

Docker can build a rootfs+application of an operating system into a container image. Sealer regards kubernetes as an operating system. The image created on this higher abstraction is a CloudImage. Realize the Build share run of the entire cluster!!!

With CloudImage, it will be extremely simple for users to practice cloud native ecological technology, such as:

Install a kubernetes cluster:

sealer run kubernetes:1.19.2 # Run a kubernetes cluster on the public cloud
sealer run kubernetes:1.19.2 --master 3 --node 3 # Run a kuberentes cluster with a specified number of nodes on the public cloud

Install to an existing machine

sealer run kuberntes:1.19.2 --master 192.168.0.2,192.168.0.3,192.168.0.4 --node 192.168.0.5,192.168.0.6,192.168.0.7

Install prometheus cluster

sealer run prometheus:2.26.0

The above command can help you install a kubernetes cluster that includes prometheus. Similarly, other software such as istio ingress grafana can be run in this way.

It's not over yet, the best thing about Sealer is that it is very convenient for users to customize a CloudImage, which is described and built through a file like Dockerfile:

Kubefile:

FROM registry.cn-qingdao.aliyuncs.com/sealer/cloudrootfs:v1.16.9-alpha.6
RUN wget https://raw.githubusercontent.com/kubernetes/dashboard/v2.2.0/aio/deploy/recommended.yaml
CMD kubectl apply -f recommended.yaml

sealer build -t registry.cn-qingdao.aliyuncs.com/sealer/dashboard:latest.

Then a CloudImage containing the dashboard is created, which can be run or shared with others.

Push the created CloudImage to the registry, which is compatible with the docker image registry, so you can push the CloudImage to docker hub, Ali ACR, or Harbor

sealer push registry.cn-qingdao.aliyuncs.com/sealer/dashboard:latest

• An extremely simple way to install kubernetes and other software in the kubernetes ecosystem in a production or offline environment.
• Through Kubefile, you can easily customize the kubernetes CloudImage to package the cluster and applications, and submit them to the registry.
• Powerful life cycle management capabilities, to perform operations such as cluster upgrade, cluster backup and recovery, node expansion and contraction in unimaginable simple ways
• Very fast, complete cluster installation within 3 minutes
• Support ARM x86, v1.20 and above versions support containerd, almost compatible with all Linux operating systems that support systemd
• Does not rely on ansible haproxy keepalived, high availability is achieved through ipvs, takes up less resources, is stable and reliable
• There are very few in the official warehouse. Many ecological software images can be used directly, including all dependencies, one-click installation

Install a kubernetes cluster

sealer run kubernetes:v1.19.2 --master 192.168.0.2

If it is installed on the cloud:

export ACCESSKEYID=xxx
export ACCESSKEYSECRET=xxx
sealer run registry.cn-qingdao.aliyuncs.com/sealer/dashboard:latest

Or specify the number of nodes to run the cluster

sealer run registry.cn-qingdao.aliyuncs.com/sealer/dashboard:latest \
  --masters 3 --nodes 3

[root@iZm5e42unzb79kod55hehvZ ~]# kubectl get node
NAME                    STATUS ROLES AGE VERSION
izm5e42unzb79kod55hehvz Ready master 18h v1.16.9
izm5ehdjw3kru84f0kq7r7z Ready master 18h v1.16.9
izm5ehdjw3kru84f0kq7r8z Ready master 18h v1.16.9
izm5ehdjw3kru84f0kq7r9z Ready <none> 18h v1.16.9
izm5ehdjw3kru84f0kq7raz Ready <none> 18h v1.16.9
izm5ehdjw3kru84f0kq7rbz Ready <none> 18h v1.16.9

View the default startup configuration of the CloudImage:

sealer config registry.cn-qingdao.aliyuncs.com/sealer/dashboard:latest

Use Clusterfile to pull up a k8s cluster Use the provided official basic image (sealer/cloudrootfs:v1.16.9-alpha.6) to quickly pull up a k8s cluster.

Clusterfile content:

apiVersion: sealer.aliyun.com/v1alpha1
kind: Cluster
metadata:
  name: my-cluster
spec:
  image: registry.cn-qingdao.aliyuncs.com/sealer/cloudrootfs:v1.16.9-alpha.5
  provider: BAREMETAL
  ssh:
    passwd:
    pk: xxx
    pkPasswd: xxx
    user: root
  network:
    interface: eth0
    cniName: calico
    podCIDR: 100.64.0.0/10
    svcCIDR: 10.96.0.0/22
    withoutCNI: false
  certSANS:
    -aliyun-inc.com
    -10.0.0.2
    
  masters:
    ipList:
     -172.20.125.234
     -172.20.126.5
     -172.20.126.6
  nodes:
    ipList:
     -172.20.126.8
     -172.20.126.9
     -172.20.126.10

[root@iZm5e42unzb79kod55hehvZ ~]# sealer apply -f Clusterfile
[root@iZm5e42unzb79kod55hehvZ ~]# kubectl get node

apiVersion: sealer.aliyun.com/v1alpha1
kind: Cluster
metadata:
  name: my-cluster
spec:
  image: registry.cn-qingdao.aliyuncs.com/sealer/cloudrootfs:v1.16.9-alpha.5
  provider: ALI_CLOUD
  ssh:
    passwd:
    pk: xxx
    pkPasswd: xxx
    user: root
  network:
    interface: eth0
    cniName: calico
    podCIDR: 100.64.0.0/10
    svcCIDR: 10.96.0.0/22
    withoutCNI: false
  certSANS:
    -aliyun-inc.com
    -10.0.0.2
    
  masters:
    cpu: 4
    memory: 4
    count: 3
    systemDisk: 100
    dataDisks:
    -100
  nodes:
    cpu: 4
    memory: 4
    count: 3
    systemDisk: 100
    dataDisks:
    -100

Some source information of the basic settings will be written to the Clusterfile and stored in /root/.sealer/[cluster-name]/Clusterfile, so the cluster can be released as follows:

sealer delete -f /root/.sealer/my-cluster/Clusterfile