centos 7 yum安装kubernetes cluster 1.0
在这里，我会展示如何安装一个kubernetes cluster,包含1个master 2个minions。
环境需求：
centos7 64位系统 三台机器
master：192.168.5.131
minions:192.168.5.132
minions:192.168.5.133
kubernetes 的组件：
etcd
flannel
kube-apiserver
kube-controller-manager
kube-scheduler
kubelet
kube-proxy
一、部署到centos7
图以后再配
先决条件
-
每台机器禁用iptables 避免和docker 的iptables冲突：
|
1
2
|
$ systemctl stop firewalld
$ systemctl disable firewalld
|
2. 安装NTP并确保正常运行
|
1
2
3
|
$ yum -y
install
ntp
$ systemctl start ntpd
$ systemctl
enable
ntpd
|
3. 2个minions机器安装docker
|
1
2
3
|
yum
install
docker -y
yum update -y
reboot
|
CentOS系统，使用devicemapper作为存储后端，初始安装docker 会使用loopback, 导致docker启动报错。需要update之后再启动
ps aux|grep docker 结果如下
|
1
|
/usr/bin/docker
-d --selinux-enabled --storage-driver devicemapper --storage-opt dm.fs=xfs --storage-opt dm.thinpooldev=
/dev/mapper/VolGroup00-docker--pool
|
二、安装kubernetes master
以下步骤均在master上执行
-
通过yum安装etcd和kubernetes
|
1
|
yum -y
install
etcd kubernetes
|
2.修改配置文件/etc/etcd/etcd.conf,确保etcd监听所有地址，修改如下：
|
1
2
3
|
ETCD_NAME=default
ETCD_DATA_DIR=
"/var/lib/etcd/default.etcd"
ETCD_LISTEN_CLIENT_URLS=
"http://0.0.0.0:2379"
|
3.修改配置文件/etc/kubernetes/apiserver,修改如下：
|
1
2
3
4
5
6
7
|
KUBE_API_ADDRESS=
"--address=0.0.0.0"
KUBE_API_PORT=
"--port=8080"
KUBELET_PORT=
"--kubelet_port=10250"
KUBE_ETCD_SERVERS=
"--etcd_servers=http://127.0.0.1:2379"
KUBE_SERVICE_ADDRESSES=
"--portal_net=10.254.0.0/16"
KUBE_ADMISSION_CONTROL=
"--admission_control=NamespaceAutoProvision,LimitRanger,ResourceQuota"
KUBE_API_ARGS=
""
|
4.修改配置文件/etc/kubernetes/controller-manager，定义minions ip地址
|
1
|
KUBELET_ADDRESSES=
"--machines=192.168.5.132,192.168.5.133"
|
5.启动服务
|
1
2
3
4
5
|
for
SERVICES
in
etcd kube-apiserver kube-controller-manager kube-scheduler;
do
systemctl restart $SERVICES
systemctl
enable
$SERVICES
systemctl status $SERVICES
done
|
6.定义flannel网络配置到etcd,这个配置会推送到各个minions的flannel服务上
|
1
|
etcdctl mk
/coreos
.com
/network/config
'{"Network":"172.17.0.0/16"}'
|
三、安装minions
以下操作均在minions1,minions2上执行
-
安装flannel和kubernetes
|
1
|
yum -y
install
flannel kubernetes
|
2. 为etcd服务配置flannel，修改配置文件 /etc/sysconfig/flanneld
|
1
|
FLANNEL_ETCD=
"http://192.168.5.131:2379"
|
3.修改kubernetes配置文件，指定master。/etc/kubernetes/config
|
1
|
KUBE_MASTER=
"--master=http://192.168.5.131:8080"
|
4.配置kubelet服务。/etc/kubernetes/kubelet
minions1:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
###
# kubernetes kubelet (minion) config
# The address for the info server to serve on (set to 0.0.0.0 or "" for all interfaces)
KUBELET_ADDRESS=
"--address=192.168.5.132"
# The port for the info server to serve on
KUBELET_PORT=
"--port=10250"
# You may leave this blank to use the actual hostname
KUBELET_HOSTNAME=
"--hostname_override=192.168.5.132"
# location of the api-server
KUBELET_API_SERVER=
"--api_servers=http://192.168.5.131:8080"
# Add your own!
KUBELET_ARGS=
""
|
minions2:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
###
# kubernetes kubelet (minion) config
# The address for the info server to serve on (set to 0.0.0.0 or "" for all interfaces)
KUBELET_ADDRESS=
"--address=192.168.5.133"
# The port for the info server to serve on
KUBELET_PORT=
"--port=10250"
# You may leave this blank to use the actual hostname
KUBELET_HOSTNAME=
"--hostname_override=192.168.5.133"
# location of the api-server
KUBELET_API_SERVER=
"--api_servers=http://192.168.5.131:8080"
# Add your own!
KUBELET_ARGS=
""
|
5.启动服务
|
1
2
3
4
5
|
for
SERVICES
in
kube-proxy kubelet docker flanneld;
do
systemctl restart $SERVICES
systemctl
enable
$SERVICES
systemctl status $SERVICES
done
|
6.在每个minions可以看到2块网卡：docker0和flannel0，这2块网卡的ip在不同的机器ip地址不同
minion1:
|
1
2
|
ip a |
grep
flannel |
grep
inet
inet 172.17.20.0
/16
scope global flannel0
|
minion2:
|
1
2
|
$ ip a |
grep
flannel |
grep
inet
inet 172.17.21.0
/16
scope global flannel0
|
7.现在登陆master，确认minions的状态
|
1
2
3
4
|
[root@k8s_master ~]
# kubectl get nodes
NAME LABELS STATUS
192.168.5.132 kubernetes.io
/hostname
=192.168.5.132 Ready
192.168.5.133 kubernetes.io
/hostname
=192.168.5.133 Ready
|
太棒了！kubernetes的集群就配置完成了，下面让我们开始弄pod
在创建pod如果有报错为
|
1
|
is forbidden: no API token found
for
service account default
/default
, retry after the token is automatically created and added to the service account
|
如果你在apiserver配置如下：
|
1
2
|
# default admission control policies#
KUBE_ADMISSION_CONTROL=
"--admission_control=NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ServiceAccount,ResourceQuota"
|
需要去掉
|
1
|
SecurityContextDeny,ServiceAccount
|
本文转自银狐博客51CTO博客，原文链接http://blog.51cto.com/foxhound/1684391如需转载请自行联系原作者
战狐