docker (centOS 7) 使用笔记4

本次测试的系统包含centos 7.2 64 bit，centos 7.3 64 bit

1. 安装

yum -y install etcd

2. 配置

此处一共准备了3台机器(10.10.10.100, 10.10.10.101, 10.10.10.102) 

(1) etcd服务的配置和启动方式

启用etcd服务

systemctl enable etcd

修改服务配置 /etc/systemd/system/multi-user.target.wants/etcd.service，其中的启动指令(ExecStart=)

修改配置后，装载配置，重启服务

systemctl daemon-reload
systemctl restart etcd

启动成功后，可以看到新增加了2个监听端口：2379, 2380

(2) 集群配置

第1个节点(10.10.10.100) (注意：第一次启动时，--initial-cluster-state new，启动成功后，集群已经创建，此时需修改为  --initial-cluster-state existing。否则下次重启服务会失败。):

ExecStart=/bin/bash -c "GOMAXPROCS=$(nproc) /usr/bin/etcd --name="etcd0" --data-dir="${ETCD_DATA_DIR}" --listen-client-urls="http://0.0.0.0:2379" --listen-peer-urls="http://0.0.0.0:2380" --advertise-client-urls="http://10.10.10.100:2379" --initial-advertise-peer-urls="http://10.10.10.100:2380" --initial-cluster-token="etcd-cluster-1" --initial-cluster="etcd0=http://10.10.10.100:2380,etcd1=http://10.10.10.101:2380,etcd2=http://10.10.10.102:2380" --initial-cluster-state new " 

后面的节点可以先手工添加信息:

etcdctl member add etcd1 http://10.10.10.101:2380
etcdctl member add etcd2 http://10.10.10.102:2380  

第2个节点(10.10.10.101):

ExecStart=/bin/bash -c "GOMAXPROCS=$(nproc) /usr/bin/etcd --name="etcd1" --data-dir="${ETCD_DATA_DIR}" --listen-client-urls="http://0.0.0.0:2379" --listen-peer-urls="http://0.0.0.0:2380" --advertise-client-urls="http://10.10.10.101:2379" --initial-advertise-peer-urls="http://10.10.10.101:2380" --initial-cluster-token="etcd-cluster-1" --initial-cluster="etcd0=http://10.10.10.100:2380,etcd1=http://10.10.10.101:2380,etcd2=http://10.10.10.102:2380"  " 

第3个节点(10.10.10.102):

ExecStart=/bin/bash -c "GOMAXPROCS=$(nproc) /usr/bin/etcd --name="etcd2" --data-dir="${ETCD_DATA_DIR}" --listen-client-urls="http://0.0.0.0:2379" --listen-peer-urls="http://0.0.0.0:2380" --advertise-client-urls="http://10.10.10.102:2379" --initial-advertise-peer-urls="http://10.10.10.102:2380" --initial-cluster-token="etcd-cluster-1" --initial-cluster="etcd0=http://10.10.10.100:2380,etcd1=http://10.10.10.101:2380,etcd2=http://10.10.10.102:2380" " 

启动成功后可以查看集群状态，member 列表

# etcdctl cluster-health
member 17294cb126466d4d is healthy: got healthy result from http://10.10.10.100:2379
member a17abe451cf50cbd is healthy: got healthy result from http://10.10.10.101:2379
member f4a143b3a5f1cdf7 is healthy: got healthy result from http://10.10.10.102:2379
cluster is healthy

# etcdctl member list
17294cb126466d4d: name=etcd2 peerURLs=http://10.10.10.100:2380 clientURLs=http://10.28.148.46:2379 isLeader=false
a17abe451cf50cbd: name=etcd0 peerURLs=http://10.10.10.101:2380 clientURLs=http://10.28.148.61:2379 isLeader=true
f4a143b3a5f1cdf7: name=etcd1 peerURLs=http://10.10.10.102:2380 clientURLs=http://10.28.148.57:2379 isLeader=false