K8s/k8s/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S/k8S
This article is in the group @Tianrain-Xiaowei @HongXiaolu @Logan and other gods under the guidance, and according to the “native Accelerated Chinese Kubernetes installation”, the final solution
Environment to prepare
| The host name | Host IP | OS | docker version | ranhcer version |
|---|---|---|---|---|
| anjia-ubuntu | 192.168.31.83 | Ubuntu 17.04 4.9.0-12 – generic x86_64 | Docker version 1.12.6 | V1.6.11 – rc10 |
Install the docker
Docker-ce VERSIONS Getting Started with Hosts#SUPPORTED DOCKER VERSIONS install the SUPPORTED docker-ce version. Tencent Cloud image, Tsinghua Image, etc.)
Rancher K8s only supports Docker 1.12.3+ at present
sudo apt install docker.ioCopy the codeInstall the rancher
According to the actual situation of Installing Rancher Server, it is recommended to use accelerator DaoCloud – the leading container cloud platform or Ali Cloud Docker accelerator
If rancher/ Server is v1.6.10 (not tried below V1.6.10), you need to modify the private Registry and push the gCR. IO plug-in to the private Registry, The namespace must be google_containers. If allowed, select V1.6.11 +(currently V1.6.11 is rc version). Select the version according to the actual situation
sudo docker run -d --restart=unless-stopped --name=rancher-server -p 8080:8080 rancher/server:v1.6.11-rc10 && sudo docker logs -f rancher-server
#sudo docker run -d --restart=unless-stopped --name=rancher-server -p 8080:8080 rancher/server:v1.6.10 && sudo docker logs -f rancher-serverCopy the coderegistereddocker hub
Install k8s
If docker and K8S have been installed before, run this command
docker rm -f -v $(docker ps -aq)
docker volume rm $(docker volume ls)
sudo rm -rf /var/etcd/Copy the codeCreating an Environment Template
Example Modify a K8S template
Private Registry for Add-Ons and Pod Infra Container Image index.docker.io
Image namespace for Add-Ons and Pod Infra Container Image anjia0532
Image namespace for kubernetes-helm Image anjia0532
Pod Infra Container Image anjia0532
Create the K8S environment
Select the K8S environment and add hosts
View k8S basic service status
When the base service is green, it can be used
View the K8S Dashboard
Abnormal screening
If you open dashboard report 503 ServiceUnavailable, thank you very much @Tianra-Li Xiaowei in the group for patiently answering this question and solving several problems by referring to the top 10 most common reasons for Kubernetes deployment failures (Part 1)
Open the Cli
> kubectl --namespace=kube-system get pods
NAME READY STATUS RESTARTS AGE
heapster-79684d56d6-8pjrd 1/1 Running 0 13m
kube-dns-7f59fd996-nkvv5 3/3 Running 0 13m
kubernetes-dashboard-86d9cc5b4-7lxj5 0/1 ImagePullBackOff 0 13m
monitoring-grafana-6dc7576774-8x79x 1/1 Running 0 13m
monitoring-influxdb-d78f84c6c-29wcp 1/1 Running 0 13m
tiller-deploy-c4598db7d-8wxpp 1/1 Running 0 13m
# after
> kubectl --namespace=kube-system describe pod kubernetes-dashboard-86d9cc5b4-7lxj5
# this is normal Running log,ImagePullBackOff did not cut down
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal Scheduled 16m default-scheduler Successfully assigned kubernetes-dashboard-86d9cc5b4-7lxj5 to k8s
Normal SuccessfulMountVolume 16m kubelet, k8s MountVolume.SetUp succeeded for volume "io-rancher-system-token-lb68r"
Normal Pulled 16m kubelet, k8s Container image "Index. The docker. IO/anjia0532 / kubernetes - dashboard - amd64: v1.7.1" already present on machine
Normal Created 16m kubelet, k8s Created container
Normal Started 16m kubelet, k8s Started container
# You can also help troubleshoot problems according to events
> kubectl --namespace=kube-system get eventsCopy the codeBlog anjia. Ml / 2017/11/13 /… The nuggets juejin. Cn/post / 684490… Jane books www.jianshu.com/p/2f906a7f4…