3.4 部署controller-manager
3.4.1 集群规划
| 主机名 | 角色 | IP地址 |
|---|---|---|
| pg60-21.k8s.host.com | controller-manager | 10.20.60.21 |
| pg60-22.k8s.host.com | controller-manager | 10.20.60.22 |
| pg60-23.k8s.host.com | controller-manager | 10.20.60.23 |
3.4.2 创建启动脚本
在 pg60-21.k8s.host.com 虚机实例上操作，另外一台运算节点安装部署方法类似。
shell> cat /opt/kubernetes/server/sbin/kube-controller-manager.sh
#!/bin/sh
/opt/kubernetes/server/bin/kube-controller-manager \
--cluster-cidr 172.60.0.0/16 \
--leader-elect true \
--log-dir /export/kubernetes/logs/kube-controller-manager \
--master http://127.0.0.1:8080 \
--service-account-private-key-file /opt/kubernetes/server/ssl/ca-key.pem \
--service-cluster-ip-range 192.168.0.0/16 \
--root-ca-file /opt/kubernetes/server/ssl/ca.pem \
--v 2
shell> chmod +x /opt/kubernetes/server/sbin/kube-controller-manager.sh
shell> mkdir -p /export/kubernetes/logs/kube-controller-manager
3.4.3 在supervisor中创建kube-controller-manager.conf的配置
shell> cat > /opt/supervisor/conf.d/kube-controller-manager.conf << EOF
[program:kube-controller-manager-60-21]
command=/opt/kubernetes/server/sbin/kube-controller-manager.sh
numprocs=1
directory=/opt/kubernetes/server/bin
autostart=true
autorestart=true
startsecs=22
startretries=3
exitcodes=0,2
stopsignal=QUIT
stopwaitsecs=10
user=root
redirect_stderr=false
stdout_logfile=/export/kubernetes/logs/kube-controller-manager/controll.stdout.log
stdout_logfile_maxbytes=64MB
stdout_logfile_backups=4
stdout_capture_maxbytes=1MB
stdout_events_enabled=false
stderr_logfile=/export/kubernetes/logs/kube-controller-manager/controll.stderr.log
stderr_logfile_maxbytes=64MB
stderr_logfile_backups=4
stderr_capture_maxbytes=1MB
stderr_events_enabled=false
EOF
3.4.4 启动kube-controller-manager服务并检查
shell> supervisorctl update
kube-controller-manager-60-21: added process group
shell> supervisorctl status
etcd-server-60-21 RUNNING pid 1755, uptime 0:04:39
kube-apiserver-60-21 RUNNING pid 1756, uptime 0:04:39
kube-controller-manager-60-21 RUNNING pid 1832, uptime 0:00:25