一、什么是consul
1、Consul 是 HashiCorp 公司推出的开源软件,用于实现分布式系统的服务发现与配置。
Consul 是分布式的、高可用的、 可横向扩展的
2、官方网站:
https://www.consul.io/
3、Consul 集群间使用了 Gossip 协议通信和 raft 一致性算法
二、下载consul软件
1,下载地址:
https://www.consul.io/downloads
选择64位linux版本下载
2,把consul的安装文件解压:
[root@localhost consul]# cd /usr/local/source/consul
[root@localhost consul]# ls
consul_1.8.4_linux_amd64.zip
[root@localhost consul]# unzip consul_1.8.4_linux_amd64.zip
Archive: consul_1.8.4_linux_amd64.zip
inflating: consul
[root@localhost consul]# ls
consul consul_1.8.4_linux_amd64.zip
3,复制到到各服务器的/usr/local/soft目录下
三、在第一台consul服务器上运行:
1,生成数据目录:
[root@consul1 /]# mkdir /data/
[root@consul1 /]# mkdir /data/consul/
[root@consul1 /]# mkdir /data/consul/data
[root@consul1 /]# chmod 777 /data/consul/data
2,运行consul
-server:以server身份启动
-bootstrap-expect=2:集群要求的最少server数量
-bind:监听的ip
-client:客户端ip,0.0.0.0表示不限制客户端ip
-data-dir:指定存放数据的目录
-node:指定节点id,注意:同一集群内节点id不允许重复
[root@consul1 /]# nohup /usr/local/soft/consul agent -server -bind=172.17.0.2 -client=0.0.0.0 -bootstrap-expect=2 -data-dir=/data/consul/da -node=server-2 >/dev/null 2>&1 &
[1] 549
[root@consul1 /]#
查看是否在运行中,这里我们选择查看端口:
[root@consul1 /]# ss -lntp
State Recv-Q Send-Q Local Address:Port Peer Address:Port
LISTEN 0 4096 172.17.0.2:8300 0.0.0.0:* users:(("consul",pid=549,fd=6))
LISTEN 0 4096 172.17.0.2:8301 0.0.0.0:* users:(("consul",pid=549,fd=12))
LISTEN 0 4096 172.17.0.2:8302 0.0.0.0:* users:(("consul",pid=549,fd=8))
LISTEN 0 4096 *:8500 *:* users:(("consul",pid=549,fd=16))
LISTEN 0 4096 *:8600 *:* users:(("consul",pid=549,fd=15))
可以看到consul已经在守护端口中,而且consul启用了多个端口
3,查看consul的版本:
[root@consul1 /]# /usr/local/soft/consul --version
Consul v1.8.4
Revision 12b16df32
Protocol 2 spoken by default, understands 2 to 3 (agent will automatically use protocol >2 when speaking to compatible agents)
4,查看consul的集群成员数量
[root@consul1 /]# /usr/local/soft/consul members
Node Address Status Type Build Protocol DC Segment
server-2 172.17.0.2:8301 alive server 1.8.4 2 dc1 <all>
只有一台机器,正常
5,查看当前节点的信息:
[root@consul1 /]# /usr/local/soft/consul info
agent:
check_monitors = 0
check_ttls = 0
checks = 0
services = 0
build:
prerelease =
revision = 12b16df3
version = 1.8.4
consul:
acl = disabled
bootstrap = false
known_datacenters = 1
leader = false
leader_addr =
server = true
raft:
applied_index = 0
commit_index = 0
fsm_pending = 0
last_contact = never
last_log_index = 0
last_log_term = 0
last_snapshot_index = 0
last_snapshot_term = 0
latest_configuration = []
latest_configuration_index = 0
num_peers = 0
protocol_version = 3
protocol_version_max = 3
protocol_version_min = 0
snapshot_version_max = 1
snapshot_version_min = 0
state = Follower
term = 0
runtime:
arch = amd64
cpu_count = 2
goroutines = 79
max_procs = 2
os = linux
version = go1.14.6
serf_lan:
coordinate_resets = 0
encrypted = false
event_queue = 0
event_time = 1
failed = 0
health_score = 0
intent_queue = 0
left = 0
member_time = 1
members = 1
query_queue = 0
query_time = 1
serf_wan:
coordinate_resets = 0
encrypted = false
event_queue = 0
event_time = 1
failed = 0
health_score = 0
intent_queue = 0
left = 0
member_time = 1
members = 1
query_queue = 0
query_time = 1
四、在第二台consul服务器上运行相同的操作,
启动的命令需要修改ip和节点名:
[root@consul2 /]# nohup /usr/local/soft/consul agent -server -bind=172.17.0.3 -client=0.0.0.0 -bootstrap-expect=2 -data-dir=/data/consul/da -node=server-3 >/dev/null 2>&1 &
[1] 371
查看集群成员:
[root@consul2 /]# /usr/local/soft/consul members
Node Address Status Type Build Protocol DC Segment
server-3 172.17.0.3:8301 alive server 1.8.4 2 dc1 <all>
把当前节点加入到第一台consul服务器的ip:
[root@consul2 /]# /usr/local/soft/consul join 172.17.0.2
Successfully joined cluster by contacting 1 nodes.
加入成功后再次查看节点:
[root@consul2 /]# /usr/local/soft/consul members
Node Address Status Type Build Protocol DC Segment
server-2 172.17.0.2:8301 alive server 1.8.4 2 dc1 <all>
server-3 172.17.0.3:8301 alive server 1.8.4 2 dc1 <all>
五、在第三台consul服务器上运行相同的操作:
命令修改ip和节点名:并允许访问ui
-ui:允许访问web ui
[root@consul3 /]# nohup /usr/local/soft/consul agent -server -bind=172.17.0.4 -client=0.0.0.0 -bootstrap-expect=2 -data-dir=/data/consul/da -node=server-4 -ui >/dev/null 2>&1 &
[1] 229
[root@consul3 /]#
查看成员
[root@consul3 /]# /usr/local/soft/consul members
Node Address Status Type Build Protocol DC Segment
server-4 172.17.0.4:8301 alive server 1.8.4 2 dc1 <all>
加入集群
[root@consul3 /]# /usr/local/soft/consul join 172.17.0.2
Successfully joined cluster by contacting 1 nodes.
成功后再次查看成员
[root@consul3 /]# /usr/local/soft/consul members
Node Address Status Type Build Protocol DC Segment
server-2 172.17.0.2:8301 alive server 1.8.4 2 dc1 <all>
server-3 172.17.0.3:8301 alive server 1.8.4 2 dc1 <all>
server-4 172.17.0.4:8301 alive server 1.8.4 2 dc1 <all>
六,访问在第三台consul服务器上启用的web ui:
访问:
http://172.17.0.4:8500/
返回:
点击后可以查看集群内3个实例:
如图:
点击每台机器可查看其状态:
七,consul的退出:
leave指令触发一个优雅的离开动作并关闭agent,节点离开后不会尝试重新加入集群中
在第二台consul服务器上执行:
[root@consul2 /]# /usr/local/soft/consul leave
Graceful leave complete
查看端口:
[root@consul2 /]# ss -lntp
State Recv-Q Send-Q Local Address:Port Peer Address:Port
可以看到守护进程已退出
在其他节点查看成员:
[root@consul3 /]# /usr/local/soft/consul members
Node Address Status Type Build Protocol DC Segment
server-2 172.17.0.2:8301 alive server 1.8.4 2 dc1 <all>
server-3 172.17.0.3:8301 left server 1.8.4 2 dc1 <all>
server-4 172.17.0.4:8301 alive server 1.8.4 2 dc1 <all>
可以看到server-3的状态已变更为left
从web ui查看:
已看不到 server-3
八、docker安装consul
docker run -d -p 8500:8500 –restart=always –name consul-8500 consul:1.8.8 agent -server -bootstrap -ui -node=consul_node_01 -client=‘0.0.0.0’
-d: 后台运行容器,并返回容器ID;
-p: 指定端口映射,格式为:主机(宿主)端口:容器端口
--name consul-8500 : 指定容器名称,自定义
consul:1.8.8 : 指定镜像, 镜像名:标签名
agent: 表示启动 Agent 进程。
server:表示启动 Consul Server 模式
client:表示启动 Consul Cilent 模式。
bootstrap:表示这个节点是 Server-Leader ,每个数据中心只能运行一台服务器。技术角度上讲 Leader 是通过 Raft 算法选举的,但是集群第一次启动时需要一个引导Leader,在引导群集后,建议不要使用此标志。
ui:表示启动 Web UI 管理器 。
-node=consul_node_01:节点的名称,自定义;集群中必须是唯一的,默认是该节点的主机名。
-client:consul服务侦听地址,这个地址提供HTTP、DNS、RPC等服务,
默认是127.0.0.1所以不对外提供服务,如果你要对外提供服务改成0.0.0.0。
join:表示加入到某一个集群中去。 如:-join=192.168.1.169。