Deploy a kubernetes cluster on CentOS7.0

Deploy a kubernetes cluster on CentOS7.0
I. Deployment environment and architecture

• OpenStack: hengtian cloud 3.4

• Operating System: centos 7.0

• Flannel: 0.5.5

• Kubernetes: 1.2.0

• Etcd version: 2.2.2

• Docker version: 1.18.0

• Cluster information:

Role

Hostname

IP Address

Master

Master

10.0.222.2

Node

Node1

10.0.222.3

Node

Node2

10.0.222.4

Node

Node3

10.0.222.5

Node

Node4

10.0.222.6

The master node contains four components: kube-apiserver kube-scheduler kube-controller-manager etcd.

Node contains two components: kube-proxy and kubelet.

1. kube-apiserver: Located on the master node, which accepts user requests.
2. kube-scheduler: Located on the master node, responsible for resource scheduling, that is, the node on which the pod is created.
3. kube-controller-manager: Located on the master node, including ReplicationManager, Endpointscontroller, Namespacecontroller, and Nodecontroller.
4. etcd: Distributed Key-value storage system that shares the resource object information of the entire cluster.
5. kubelet: it is located on a node and is responsible for maintaining the pods running on a specific host.
6. kube-proxy: located on a node. It acts as a Service proxy.

Ii. Installation Steps

Preparations

Disable Firewall

To avoid conflicts with Docker's iptables, We need to disable the firewall on the node:

12

$ systemctl stop firewalld$ systemctl disable firewalld

Install NTP

To keep the time of each server consistent, you also need to install NTP for all servers:

123

$ yum -y install ntp$ systemctl start ntpd$ systemctl enable ntpd

Deploy the Master to install etcd and kubernetes

1

$ yum -y install etcd kubernetes

Configure etcd

Modify the configuration file of etcd/etc/etcd/etcd.conf:

123

ETCD_NAME=defaultETCD_DATA_DIR="/var/lib/etcd/default.etcd"ETCD_LISTEN_CLIENT_URLS="http://0.0.0.0:2379"

Configure the network in etcd

Define the network configuration in etcd. The flannel service in nodeN pulls this configuration.

1

$ etcdctl mk /coreos.com/network/config '{"Network":"172.17.0.0/16"}'

Configure the Kubernetes API server

1234567

API_ADDRESS="--address=0.0.0.0"KUBE_API_PORT="--port=8080"KUBELET_PORT="--kubelet_port=10250"KUBE_ETCD_SERVERS="--etcd_servers=http://10.0.222.2:2379"KUBE_SERVICE_ADDRESSES="--portal_net=10.254.0.0/16"KUBE_ADMISSION_CONTROL="--admission_control=NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ResourceQuota"KUBE_API_ARGS=""

Note thatKUBE_ADMISSION_CONTROLDefault includeServiceAccountDelete it. Otherwise, an error is reported when the API server is started.

Start the service

Next, start the following services on the Master:

12345

$for SERVICES in etcd kube-apiserver kube-controller-manager kube-scheduler; do systemctl restart $SERVICESsystemctl enable $SERVICESsystemctl status $SERVICES done

Deploy Node to install Kubernetes and Flannel

1

$ yum -y install flannel kubernetes

Configure Flannel

Modify the configuration file of Flannel/etc/sysconfig/flanneld:

123

FLANNEL_ETCD="http://10.0.222.2:2379"FLANNEL_ETCD_KEY="/coreos.com/network"FLANNEL_OPTIONS="--iface=ens3"

Note thatFLANNEL_OPTIONSThe value of iface in is the network card of your own server. Different servers and configurations are different from mine.

Start Flannel

123

$systemctl restart flanneld$systemctl enable flanneld$systemctl status flanneld

Upload Network Configuration

Createconfig.json, The content is as follows:

12345678

{"Network": "172.17.0.0/16","SubnetLen": 24,"Backend": {     "Type": "vxlan",     "VNI": 7890     } }

Then upload the configuration to the etcd Server:

1

$ curl -L http://10.0.222.2:2379/v2/keys/coreos.com/network/config -XPUT --data-urlencode value@config.json

Modify Kubernetes configurations

Modify the default configuration file of kubernetes/etc/kubernetes/config:

1

KUBE_MASTER="--master=http://10.0.222.2:8080"

Modify kubelet configurations

Modify the configuration file of the kubelet Service/etc/kubernetes/kubelet:

123456

KUBELET_ADDRESS="--address=0.0.0.0"KUBELET_PORT="--port=10250"# change the hostname to minion IP addressKUBELET_HOSTNAME="--hostname_override=node1"KUBELET_API_SERVER="--api_servers=http://10.0.222.2:8080"KUBELET_ARGS=""

For different node nodes, you only need to change KUBELET_HOSTNAME to the hostname of the node.

Start node Service

12345

$ for SERVICES in kube-proxy kubelet docker; do systemctl restart $SERVICESsystemctl enable $SERVICESsystemctl status $SERVICES done

Create a snapshot and install it on other nodes (modify the corresponding hostname and KUBELET_HOSTNAME) to view the cluster nodes.

After the deployment is complete, you can run the kubectl command to view the status of the entire cluster:

1 kubectl-s "http: // 10.0.222.2: 8080" get nodes