前面几篇关于consul的文章简单的介绍了windows下安装consul以及consul作为注册中心和配置中心的简单使用，基于前面的基础，这里介绍下载linux下安装consul以及结合docker搭建consul集群，解决consul配置的数据无法保存的问题。
选择linux的版本的consul进行下载
把下载的linux下的安装包consul拷贝到linux环境里面，使用unzip进行解压：
如果linux下面没有unzip命令，则使用yum install unzip命令进行安装
配置如下：
export JAVA_HOME=/usr/local/jdk1.8.0_172 export MAVEN_HOME=/usr/local/apache-maven-3.5.4 export CONSUL_HOME=/usr/local/consul export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar export PATH=$JAVA_HOME/bin:$MAVEN_HOME/bin:$CONSUL_HOME:$PATH上面的CONSUL_HOME就是consul的路径，上面的配置仅供参考。
进行了配置以后，退出保存修改，使用下面的命令使配置生效：
source /etc/profile这样进行配置以后，我们就可以方便在任何地方使用consul命令了。
输出如上所示，http端口为8500，dns端口为8600，绑定的本地ip为127.0.0.1，如果consul需要端口需要被外部访问需要开发8500端口和8600端口，可以参照：
centos查询端口是不是开放的
firewall-cmd --permanent --query-port=8500/tcp#添加对外开放端口
firewall-cmd --permanent --add-port=8500/tcp#重启防火墙
firewall-cmd --reload参考：参考链接
consul agent常用命令解读 -data-dir ： 作用：指定agent储存状态的数据目录，这是所有agent都必须的,对server尤其重要，因为他们必须持久化集群的状态
-config-dir ： 作用：指定service的配置文件和检查定义所在的位置。目录必需为consul.d，文件内容都是json格式的数据。配置详解见官方
-config-file 作用：指定一个要装载的配置文件
-dev ： 作用：开发服务器模式,虽然是server模式，但不用于生产环境，因为不会有任何持久化操作，即不会有任何数据写入到磁盘
-bootstrap-expect 作用： 参数表明该服务运行时最低开始进行选举的节点数，当设置为1时，则意味允许节点为一个时也进行选举；当设置为3时，则等到3台节点同时运行consul并加入到server才能参与选举，选举完集群才能够正常工作。 一般建议服务器结点3-5个。
-node ： 作用：指定节点在集群中的名称，该名称在集群中必须是唯一的（默认这是机器的主机名），直接采用机器的IP
-bind ： 作用：指明节点的IP地址，一般是0.0.0.0或者云服务器内网地址，不能写阿里云外网地址。这是Consul侦听的地址，它必须可以被集群中的所有其他节点访问。虽然绑定地址不是绝对必要的，但最好提供一个。
-server ： 作用：指定节点为server,每个数据中心（DC）的server数推荐3-5个。
-client ： 作用：指定节点为client，指定客户端接口的绑定地址，包括：HTTP、DNS、RPC 默认是127.0.0.1，只允许回环接口访问
-datacenter ： 作用：指定机器加入到哪一个数据中心中。老版本叫-dc，-dc已经失效
Agent： Consul集群中长时间运行的守护进程，以consul agent 命令开始启动. 在客户端和服务端模式下都可以运行，可以运行DNS或者HTTP接口， 它的主要作用是运行时检查和保持服务同步。 Client： 客户端, 无状态, 以一个极小的消耗将接口请求转发给局域网内的服务端集群. Server： 服务端, 保存配置信息, 高可用集群, 在局域网内与本地客户端通讯, 通过广域网与其他数据中心通讯. 每个数据中心的 server 数量推荐为 3 个或是 5 个. Datacenter： 数据中心，多数据中心联合工作保证数据存储安全快捷 Consensus： 一致性协议使用的是Raft Protocol RPC： 远程程序通信 Gossip： 基于 Serf 实现的 gossip 协议，负责成员、失败探测、事件广播等。通过 UDP 实现各个节点之间的消息。分为 LAN 上的和 WAN 上的两种情形。
1，参数案例
前面我们使用consul agent -dev启动的consul在云服务器上是不能被外部访问的，那么要被外部访问我们需要加参数，参照如下：
consul agent -dev -http-port 8500 -client 0.0.0.0
参数说明：
-client 0.0.0.0：表明不是绑定的不是默认的127.0.0.1地址，可以通过公网进行访问
-http-port 8500：通过该参数可以修改consul启动的http端口
[root@iZbp1dmlbagds9s70r8luxZ local]# consul members Node Address Status Type Build Protocol DC Segment iZbp1dmlbagds9s70r8luxZ 127.0.0.1:8301 alive server 1.2.1 2 dc1 <all> [root@iZbp1dmlbagds9s70r8luxZ local]#
node:节点名 Address：节点地址 Status：alive表示节点健康 Type：server运行状态是server状态 DC：dc1表示该节点属于DataCenter1
前面都是以-dev开发模式启动的consul，该模式启动的consul作为配置中心的时候，配置数据是不能保存的，不能进行持久化，需要进行持久化，则需要以服务模式启动。
[root@iZbp1dmlbagds9s70r8luxZ local]# consul agent -server -ui -bootstrap-expect=1 -data-dir=/usr/local/consul/data -node=agent-one -advertise=47.98.112.71 -bind=0.0.0.0 -client=0.0.0.0 BootstrapExpect is set to 1; this is the same as Bootstrap mode. bootstrap = true: do not enable unless necessary ==> Starting Consul agent... ==> Consul agent running! Version: 'v1.2.1' Node ID: '04b82369-8b5b-19f3-ab0d-6a82266a2110' Node name: 'agent-one' Datacenter: 'dc1' (Segment: '<all>') Server: true (Bootstrap: true) Client Addr: [0.0.0.0] (HTTP: 8500, HTTPS: -1, DNS: 8600) Cluster Addr: 47.98.112.71 (LAN: 8301, WAN: 8302) Encrypt: Gossip: false, TLS-Outgoing: false, TLS-Incoming: false ==> Log data will now stream in as it occurs: 2018/07/28 10:54:02 [INFO] raft: Initial configuration (index=1): [{Suffrage:Voter ID:04b82369-8b5b-19f3-ab0d-6a82266a2110 Address:47.98.112.71:8300}] 2018/07/28 10:54:02 [WARN] memberlist: Binding to public address without encryption! 2018/07/28 10:54:02 [INFO] serf: EventMemberJoin: agent-one.dc1 47.98.112.71 2018/07/28 10:54:02 [WARN] memberlist: Binding to public address without encryption! 2018/07/28 10:54:02 [INFO] serf: EventMemberJoin: agent-one 47.98.112.71 2018/07/28 10:54:02 [INFO] agent: Started DNS server 0.0.0.0:8600 (udp) 2018/07/28 10:54:02 [INFO] raft: Node at 47.98.112.71:8300 [Follower] entering Follower state (Leader: "") 2018/07/28 10:54:02 [INFO] consul: Adding LAN server agent-one (Addr: tcp/47.98.112.71:8300) (DC: dc1) 2018/07/28 10:54:02 [INFO] consul: Handled member-join event for server "agent-one.dc1" in area "wan" 2018/07/28 10:54:02 [WARN] agent/proxy: running as root, will not start managed proxies 2018/07/28 10:54:02 [INFO] agent: Started DNS server 0.0.0.0:8600 (tcp) 2018/07/28 10:54:02 [INFO] agent: Started HTTP server on [::]:8500 (tcp) 2018/07/28 10:54:02 [INFO] agent: started state syncer 2018/07/28 10:54:08 [WARN] raft: Heartbeat timeout from "" reached, starting election 2018/07/28 10:54:08 [INFO] raft: Node at 47.98.112.71:8300 [Candidate] entering Candidate state in term 2 2018/07/28 10:54:08 [INFO] raft: Election won. Tally: 1 2018/07/28 10:54:08 [INFO] raft: Node at 47.98.112.71:8300 [Leader] entering Leader state 2018/07/28 10:54:08 [INFO] consul: cluster leadership acquired 2018/07/28 10:54:08 [INFO] consul: New leader elected: agent-one 2018/07/28 10:54:08 [INFO] consul: member 'agent-one' joined, marking health alive 2018/07/28 10:54:08 [INFO] agent: Synced node info 2018/07/28 10:54:11 [WARN] consul: error getting server health from "agent-one": context deadline exceeded 2018/07/28 10:54:12 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:14 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:16 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:18 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:20 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:20 [WARN] consul: error getting server health from "agent-one": rpc error getting client: failed to get conn: dial tcp <nil>->47.98.112.71:8300: i/o timeout 2018/07/28 10:54:23 [WARN] consul: error getting server health from "agent-one": context deadline exceeded 2018/07/28 10:54:24 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:26 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:28 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:30 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:32 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:32 [WARN] consul: error getting server health from "agent-one": rpc error getting client: failed to get conn: dial tcp <nil>->47.98.112.71:8300: i/o timeout 2018/07/28 10:54:35 [WARN] consul: error getting server health from "agent-one": context deadline exceeded 2018/07/28 10:54:36 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:38 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:40 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:42 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:44 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:44 [WARN] consul: error getting server health from "agent-one": rpc error getting client: failed to get conn: dial tcp <nil>->47.98.112.71:8300: i/o timeout 2018/07/28 10:54:47 [WARN] consul: error getting server health from "agent-one": context deadline exceeded 2018/07/28 10:54:48 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:50 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:52 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:54 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:56 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:54:56 [WARN] consul: error getting server health from "agent-one": rpc error getting client: failed to get conn: dial tcp <nil>->47.98.112.71:8300: i/o timeout 2018/07/28 10:54:59 [WARN] consul: error getting server health from "agent-one": context deadline exceeded 2018/07/28 10:55:00 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:55:02 [WARN] consul: error getting server health from "agent-one": last request still outstanding 2018/07/28 10:55:04 [WARN] consul: error getting server health from "agent-one": last request still outstanding ^C 2018/07/28 10:55:10 [INFO] agent: Caught signal: interrupt 2018/07/28 10:55:10 [INFO] agent: Graceful shutdown disabled. Exiting 2018/07/28 10:55:10 [INFO] agent: Requesting shutdown 2018/07/28 10:55:10 [INFO] consul: shutting down server 2018/07/28 10:55:10 [WARN] serf: Shutdown without a Leave 2018/07/28 10:55:10 [WARN] serf: Shutdown without a Leave 2018/07/28 10:55:10 [INFO] manager: shutting down 2018/07/28 10:55:10 [INFO] agent: consul server down 2018/07/28 10:55:10 [INFO] agent: shutdown complete 2018/07/28 10:55:10 [INFO] agent: Stopping DNS server 0.0.0.0:8600 (tcp) 2018/07/28 10:55:10 [INFO] agent: Stopping DNS server 0.0.0.0:8600 (udp) 2018/07/28 10:55:10 [INFO] agent: Stopping HTTP server [::]:8500 (tcp) 2018/07/28 10:55:11 [WARN] agent: Timeout stopping HTTP server [::]:8500 (tcp) 2018/07/28 10:55:11 [INFO] agent: Waiting for endpoints to shut down 2018/07/28 10:55:11 [INFO] agent: Endpoints down 2018/07/28 10:55:11 [INFO] agent: Exit code: 1 [root@iZbp1dmlbagds9s70r8luxZ local]# [root@iZbp1dmlbagds9s70r8luxZ local]# consul agent -server -ui -bootstrap-expect=1 -data-dir=/usr/local/consul/data -node=agent-one -advertise=47.98.112.71 -bind=0.0.0.0 -client=0.0.0.0 BootstrapExpect is set to 1; this is the same as Bootstrap mode. bootstrap = true: do not enable unless necessary ==> Starting Consul agent... ==> Consul agent running! Version: 'v1.2.1' Node ID: '04b82369-8b5b-19f3-ab0d-6a82266a2110' Node name: 'agent-one' Datacenter: 'dc1' (Segment: '<all>') Server: true (Bootstrap: true) Client Addr: [0.0.0.0] (HTTP: 8500, HTTPS: -1, DNS: 8600) Cluster Addr: 47.98.112.71 (LAN: 8301, WAN: 8302) Encrypt: Gossip: false, TLS-Outgoing: false, TLS-Incoming: false ==> Log data will now stream in as it occurs: 2018/07/28 10:55:15 [INFO] raft: Initial configuration (index=1): [{Suffrage:Voter ID:04b82369-8b5b-19f3-ab0d-6a82266a2110 Address:47.98.112.71:8300}] 2018/07/28 10:55:15 [WARN] memberlist: Binding to public address without encryption! 2018/07/28 10:55:15 [INFO] serf: EventMemberJoin: agent-one.dc1 47.98.112.71 2018/07/28 10:55:15 [WARN] memberlist: Binding to public address without encryption! 2018/07/28 10:55:15 [INFO] serf: EventMemberJoin: agent-one 47.98.112.71 2018/07/28 10:55:15 [INFO] agent: Started DNS server 0.0.0.0:8600 (udp) 2018/07/28 10:55:15 [INFO] raft: Node at 47.98.112.71:8300 [Follower] entering Follower state (Leader: "") 2018/07/28 10:55:15 [WARN] serf: Failed to re-join any previously known node 2018/07/28 10:55:15 [WARN] serf: Failed to re-join any previously known node 2018/07/28 10:55:15 [INFO] consul: Adding LAN server agent-one (Addr: tcp/47.98.112.71:8300) (DC: dc1) 2018/07/28 10:55:15 [INFO] consul: Handled member-join event for server "agent-one.dc1" in area "wan" 2018/07/28 10:55:15 [WARN] agent/proxy: running as root, will not start managed proxies 2018/07/28 10:55:15 [INFO] agent: Started DNS server 0.0.0.0:8600 (tcp) 2018/07/28 10:55:15 [INFO] agent: Started HTTP server on [::]:8500 (tcp) 2018/07/28 10:55:15 [INFO] agent: started state syncer 2018/07/28 10:55:21 [WARN] raft: Heartbeat timeout from "" reached, starting election 2018/07/28 10:55:21 [INFO] raft: Node at 47.98.112.71:8300 [Candidate] entering Candidate state in term 3 2018/07/28 10:55:21 [INFO] raft: Election won. Tally: 1 2018/07/28 10:55:21 [INFO] raft: Node at 47.98.112.71:8300 [Leader] entering Leader state 2018/07/28 10:55:21 [INFO] consul: cluster leadership acquired 2018/07/28 10:55:21 [INFO] consul: New leader elected: agent-one 2018/07/28 10:55:21 [INFO] agent: Synced node info
启动命令：
consul agent -server -ui -bootstrap-expect=1 -data-dir=/usr/local/consul/data -node=agent-one - advertise=47.98.112.71 -bind=0.0.0.0 -client=0.0.0.0
参数说明：
-server：服务器模式 -ui:能webui展示 -bootstrap-expect：server为1时即选择server集群leader -data-dir:consul状态存储文件地址 -node：指定结点名 advertise：本地ip地址 -client:指定可访问这个服务结点的ip
上面是以服务模式启动的输出，注意需要开放8300端口，consul需要进行节点通信使用
在此查看成员信息：
[root@iZbp1dmlbagds9s70r8luxZ data]# consul members Node Address Status Type Build Protocol DC Segment agent-one 47.98.112.71:8301 alive server 1.2.1 2 dc1 <all> [root@iZbp1dmlbagds9s70r8luxZ data]#加入集群的命令：consul join xx.xx.xx.xx
使用docker容器来搭建consul集群，编写Docker compose
集群说明 1,3 server 节点（consul-server1 ~ 3）和 2 node 节点（consul-node1 ~ 2） 2,映射本地 consul/data1 ~ 3/ 目录到 Docker 容器中，避免 Consul 集群重启后数据丢失。 3,Consul web http 端口分别为 8501、8502、8503
新建：docker-compose.yml
version: '2.0' services: consul-server1: image: consul:latest hostname: "consul-server1" ports: - "8501:8500" volumes: - ./consul/data1:/consul/data command: "agent -server -bootstrap-expect 3 -ui -disable-host-node-id -client 0.0.0.0" consul-server2: image: consul:latest hostname: "consul-server2" ports: - "8502:8500" volumes: - ./consul/data2:/consul/data command: "agent -server -ui -join consul-server1 -disable-host-node-id -client 0.0.0.0" depends_on: - consul-server1 consul-server3: image: consul:latest hostname: "consul-server3" ports: - "8503:8500" volumes: - ./consul/data3:/consul/data command: "agent -server -ui -join consul-server1 -disable-host-node-id -client 0.0.0.0" depends_on: - consul-server1 consul-node1: image: consul:latest hostname: "consul-node1" command: "agent -join consul-server1 -disable-host-node-id" depends_on: - consul-server1 consul-node2: image: consul:latest hostname: "consul-node2" command: "agent -join consul-server1 -disable-host-node-id" depends_on: - consul-server1集群启动时默认以 consul-server1 为 leader，然后 server2 ~ 3 和 node1 ~ 2 加入到该集群。当 server1 出现故障下线是，server2 ~ 3 则会进行选举选出新leader。
集群操作 创建并启动集群：docker-compose up -d 停止整个集群：docker-compose stop 启动集群：docker-compose start 清除整个集群：docker-compose rm（注意：需要先停止） 访问 http://localhost:8501 http://localhost:8502 http://localhost:8503
上面搭建了consul集群了，那么怎么进行负载均衡了，使用ngxin：
设定负载均衡服务器列表：
upstream consul { server 127.0.0.1:8501; server 127.0.0.1:8502; server 127.0.0.1:8503; }服务配置：
server { listen 80; server_name consul.test.com;#服务域名，需要填写你的服务域名 location / { proxy_pass http://consul;#请求转向consul服务器列表 proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; } }参考：docker搭建consul集群参考
到这里consul的相关知识基本结束了，但是这只是consul的基本使用，扩展的还需要进行consul客户端的二次开发，比如自定义consul的服务注册与发现，consul作为配置中心存在配置数据丢失的情况，怎么进行配置信息的自动备份找回。有空的话在对这一块的扩展进行深入的学习探讨。
相关文章：
Consul1-window安装consul
Consul2-使用consul作为服务注册和发现中心
Consul3-使用consul作为配置中心
转载自《Consul4-linux安装consul以及集群搭建》
