基于Mesos+Marathon+Zookeeper的Docker容器管理平台

最新推荐文章于 2024-07-16 12:10:08 发布

tianpy5

最新推荐文章于 2024-07-16 12:10:08 发布

阅读量4.9k

点赞数 1

分类专栏： Docker 文章标签： docker mesos marathon zookeeper

本文链接：https://blog.csdn.net/tianpy5/article/details/52828800

版权

Docker 专栏收录该内容

1 篇文章 0 订阅

订阅专栏

本文介绍了如何使用Mesos、Marathon和Zookeeper搭建Docker集群环境。从准备工作到安装Docker，再到启动组件、发布应用、扩展实例和调用服务，详细阐述了整个过程，最后总结了Docker集群的优势。

摘要由CSDN通过智能技术生成

刚接触docker的时候，发现docker很像一个虚拟机，随着使用的增多，它给我们的开发工作带来了极大的方便，因此就想自己搭建一个docker集群环境，以便更深入了解docker集群的工作原理。
之前曾经尝试分别安装mesos，marathon和zookeeper，组成分布式系统，但是在安装、配置过程中遇到了各种各样的问题，特别是mesos，问题比较多。既然docker容器的使用十分方便，所以我就使用基于容器的方式来搭建docker集群环境，我也推荐大家这样做，特别方便。

1. 准备工作

1.1 三台虚拟机

192.168.56.103：ubuntu-1
192.168.56.101：ubuntu-2
192.168.56.102：ubuntu-3

1.2 应用部署

三台主机分别部署zookeeper，mesos-master，mesos-slave，marathon

hostname	application
ubuntu-1	zookeeper,mesos-master,mesos-slave,marathon
ubuntu-2	zookeeper,mesos-master,mesos-slave,marathon
ubuntu-3	zookeeper,mesos-master,mesos-slave,marathon

1.3 安装docker环境

参考docker官网ubuntu操作系统安装过程
https://docs.docker.com/engine/installation/linux/ubuntulinux/

2. 下载docker镜像

docker pull mesoscloud/zookeeper   #zookeeper
docker pull mesoscloud/mesos-master   #mesos-master
docker pull mesoscloud/mesos-slave   #mesos-slave 
docker pull mesoscloud/marathon   #marathon 
docker pull davidcaste/alpine-tomcat   #tomcat

3. 启动docker镜像

以下命令分别在三台主机运行，将ip地址和相关参数修改为对应主机的ip地址和参数

3.1 运行zookeeper

docker run -d \
-e MYID=1 \ #每台主机对应一个id，三台主机分别对应1,2,3
-e SERVERS=192.168.56.103,192.168.56.101,192.168.56.102 \ #三台主机的ip地址
--name=zookeeper \
--net=host \
--restart=always \
mesoscloud/zookeeper

3.2 运行mesos-master

docker run -d \
-e MESOS_HOSTNAME=192.168.56.103 \ #第一台主机ip地址
-e MESOS_IP=192.168.56.103 \ #第一台主机ip地址
-e MESOS_QUORUM=2 \
-e MESOS_ZK=zk://192.168.56.103:2181,192.168.56.101:2181,192.168.56.102:2181/mesos \
--name mesos-master \
--net host \
--restart=always \
mesoscloud/mesos-master

3.3 运行mesos-slave

docker run -d \
-e MESOS_HOSTNAME=192.168.56.103 \ #第一台主机ip地址
-e MESOS_IP=192.168.56.103 \ #第一台主机ip地址
-e MESOS_MASTER=zk://192.168.56.103:2181,192.168.56.101:2181,192.168.56.102:2181/mesos \
-v /sys/fs/cgroup:/sys/fs/cgroup \
-v /var/run/docker.sock:/var/run/docker.sock \
--name mesos-slave \
--net host \
--privileged \
--restart=always \
mesoscloud/mesos-slave

3.4 运行marathon

docker run -d \
-e MARATHON_HOSTNAME=192.168.56.103 \ #第一台主机ip地址
-e MARATHON_HTTPS_ADDRESS=192.168.56.103 \ #第一台主机ip地址
-e MARATHON_HTTP_ADDRESS=192.168.56.103 \ #第一台主机ip地址
-e MARATHON_MASTER=zk://192.168.56.103:2181,192.168.56.101:2181,192.168.56.102:2181/mesos \
-e MARATHON_ZK=zk://192.168.56.103:2181,192.168.56.101:2181,192.168.56.102:2181/marathon \
--name marathon \
--net host \
--restart=always \
mesoscloud/marathon

4. 通过Marathon发布应用

4.1 在第一台一台主机访问http://192.168.56.103:8080，任何一台主机的8080端口都可以访问marathon，打开marathon UI，创建应用，点击Create Application，填写应用信息，发布应用，Json模式数据如下：

{
  "id": "tomcat",
  "cmd": "/opt/tomcat/bin/catalina.sh run",
  "cpus": 1,
  "mem": 512,
  "disk": 512,
  "instances": 1,
  "container": {
    "docker": {
      "image": "davidcaste/alpine-tomcat",
      "network": "BRIDGE",
      "portMappings": [
        {
          "containerPort": 8080,
          "protocol": "tcp",
          "name": null
        }
      ],
      "parameters": []
    },
    "type": "DOCKER",
    "volumes": [
      {
        "hostPath": "/var/dockertest/logs",
        "containerPath": "/logs",
        "mode": "RW"
      }
    ]
  },
  "env": {},
  "labels": {},
  "healthChecks": []
}

4.2 也可以通过marathon提供的rest接口发布应用

curl -X POST -H "Accept: application/json" -H "Content-Type: application/json" 192.168.56.103:8080/v2/apps -d'
{
  "id": "tomcat",
  "cmd": "/opt/tomcat/bin/catalina.sh run",
  "cpus": 1,
  "mem": 512,
  "disk": 512,
  "instances": 1,
  "container": {
    "docker": {
      "image": "davidcaste/alpine-tomcat",
      "network": "BRIDGE",
      "portMappings": [
        {
          "containerPort": 8080,
          "protocol": "tcp",
          "name": null
        }
      ],
      "parameters": []
    },
    "type": "DOCKER",
    "volumes": [
      {
        "hostPath": "/var/dockertest/logs",
        "containerPath": "/logs",
        "mode": "RW"
      }
    ]
  },
  "env": {},
  "labels": {},
  "healthChecks": []
}'

5. 查看应用状态

5.1 marathon主页面（http://192.168.56.103:8080）

Marathon主页面

5.2 击进入应用页面，只有一个实例

进入应用页面

5.3 点击Scale Application扩展三个实例

扩展三个实例

5.4 打开mesos主页面（http://192.168.56.103:5050），可以看到有三个tomcat实例

mesos主页面

5.5 在Mesos中打开Framework页面，我们可以看到marathon框架的正在运行。

Mesos-Framework页面

5.6 接下来我们看一下mesos-slave页面，看到运行正常的三个mesos-slave。

mesos-slave页面

6. 调用服务

我们通过marathon把tomcat运行起来了，但是tomcat里没有war包，所以我们将准备好的一个war包放进三台主机的tomcat里，这里是为了测试方便，我们直接使用docker cp命令将war包复制到docker容器当中。在实际应用中我们可以通过其他方式来加载war包，比如，在CMD命令里通过wget方式将war包下载到容器里，或者通过卷Volume的方式共享，将war包放到tomcat的webaps目录下。

suser@ubuntu-1:~$ docker ps
CONTAINER ID        IMAGE                      COMMAND                  CREATED             STATUS              PORTS                     NAMES
dd014c78d8ec        davidcaste/alpine-tomcat   "/bin/sh -c '/opt/tom"   17 minutes ago      Up 17 minutes       0.0.0.0:31204->8080/tcp   mesos-81974b67-e403-47a0-bffd-7cbfa03da51b-S0.e9647bd1-2771-487b-a01e-6a0ae68c62f9
601dbe2256d5        mesoscloud/marathon        "/usr/local/bin/dumb-"   5 hours ago         Up 5 hours                                    marathon
1dfaabf2cd3e        mesoscloud/mesos-slave     "/usr/local/bin/dumb-"   5 hours ago         Up 5 hours                                    mesos-slave
e94c3202f529        mesoscloud/mesos-master    "/usr/local/bin/dumb-"   5 hours ago         Up 5 hours                                    mesos-master
a3389a386a11        mesoscloud/zookeeper       "/usr/local/bin/dumb-"   5 hours ago         Up 5 hours                                    zookeeper
suser@ubuntu-1:~$ docker cp /mnt/share/backendresource.war dd014c78d8ec:/opt/tomcat/webapps #dd014c78d8ec为Container ID
suser@ubuntu-1:~$

接下来我们访问服务查看返回结果

suser@ubuntu-1:~$ curl 192.168.56.102:31515/backendresource/11/1/1
{"ACT_TAG":"1","X_RESULTCODE":"0"}
suser@ubuntu-1:~$ curl 192.168.56.101:31299/backendresource/11/1/1
{"ACT_TAG":"1","X_RESULTCODE":"0"}
suser@ubuntu-1:~$ curl 192.168.56.103:31204/backendresource/11/1/1
{"ACT_TAG":"1","X_RESULTCODE":"0"}

查看docker容器日志

suser@ubuntu-1:~$ docker logs dd014c78d8ec
log4j:WARN No such property [encoding] in org.apache.log4j.net.SyslogAppender.
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
2016-10-15 21:46:27 [INFO ](BackendResources              ) [TxId :  , SpanId : ] 榛戝悕鍗昩ackendresource缁撴灉:{"ACT_TAG":"1","X_RESULTCODE":"0"},鏈璋冪敤鑰楁椂:1117ms
2016-10-15 21:47:13 [INFO ](BackendResources              ) [TxId :  , SpanId : ] 榛戝悕鍗昩ackendresource缁撴灉:{"ACT_TAG":"0","X_RESULTCODE":"0"},鏈璋冪敤鑰楁椂:2ms
suser@ubuntu-1:~$

在容器也看到了调用日志，看到中文是乱码，这个可能是容器字符编码问题。

7. 通过marathon rest接口获取数据

#通过marathon接口获取所有应用信息
suser@ubuntu-1:~$ curl 192.168.56.103:8080/v2/apps
{"apps":[{"id":"/tomcat","cmd":"/opt/tomcat/bin/catalina.sh run","args":null,"user":null,"env":{},"instances":3,"cpus":0.5,"mem":512,"disk":512,"executor":"","constraints":[],"uris":[],"fetch":[],"storeUrls":[],"ports":[10000],"portDefinitions":[{"port":10000,"protocol":"tcp","labels":{}}],"requirePorts":false,"backoffSeconds":1,"backoffFactor":1.15,"maxLaunchDelaySeconds":3600,"container":{"type":"DOCKER","volumes":[],"docker":{"image":"davidcaste/alpine-tomcat","network":"BRIDGE","portMappings":[{"containerPort":8080,"hostPort":0,"servicePort":10000,"protocol":"tcp","labels":{}}],"privileged":false,"parameters":[],"forcePullImage":false}},"healthChecks":[],"readinessChecks":[],"dependencies":[],"upgradeStrategy":{"minimumHealthCapacity":1,"maximumOverCapacity":1},"labels":{},"acceptedResourceRoles":null,"ipAddress":null,"version":"2016-10-15T21:20:55.434Z","residency":null,"versionInfo":{"lastScalingAt":"2016-10-15T21:20:55.434Z","lastConfigChangeAt":"2016-10-15T16:58:42.119Z"},"tasksStaged":0,"tasksRunning":3,"tasksHealthy":0,"tasksUnhealthy":0,"deployments":[]}]}suser@ubuntu-1:~$ 

##通过marathon接口获取tomcat应用信息
suser@ubuntu-1:~$ curl 192.168.56.103:8080/v2/apps/tomcat
{"app":{"id":"/tomcat","cmd":"/opt/tomcat/bin/catalina.sh run","args":null,"user":null,"env":{},"instances":3,"cpus":0.5,"mem":512,"disk":512,"executor":"","constraints":[],"uris":[],"fetch":[],"storeUrls":[],"ports":[10000],"portDefinitions":[{"port":10000,"protocol":"tcp","labels":{}}],"requirePorts":false,"backoffSeconds":1,"backoffFactor":1.15,"maxLaunchDelaySeconds":3600,"container":{"type":"DOCKER","volumes":[],"docker":{"image":"davidcaste/alpine-tomcat","network":"BRIDGE","portMappings":[{"containerPort":8080,"hostPort":0,"servicePort":10000,"protocol":"tcp","labels":{}}],"privileged":false,"parameters":[],"forcePullImage":false}},"healthChecks":[],"readinessChecks":[],"dependencies":[],"upgradeStrategy":{"minimumHealthCapacity":1,"maximumOverCapacity":1},"labels":{},"acceptedResourceRoles":null,"ipAddress":null,"version":"2016-10-15T21:20:55.434Z","residency":null,"versionInfo":{"lastScalingAt":"2016-10-15T21:20:55.434Z","lastConfigChangeAt":"2016-10-15T16:58:42.119Z"},"tasksStaged":0,"tasksRunning":3,"tasksHealthy":0,"tasksUnhealthy":0,"deployments":[],"tasks":[{"id":"tomcat.a143d29d-92f8-11e6-b002-9e1a99d79ec2","slaveId":"81974b67-e403-47a0-bffd-7cbfa03da51b-S2","host":"192.168.56.102","startedAt":"2016-10-15T16:58:48.013Z","stagedAt":"2016-10-15T16:58:43.876Z","ports":[31515],"version":"2016-10-15T16:58:42.119Z","ipAddresses":[{"ipAddress":"172.17.0.2","protocol":"IPv4"}],"appId":"/tomcat"},{"id":"tomcat.424d101e-931d-11e6-b002-9e1a99d79ec2","slaveId":"81974b67-e403-47a0-bffd-7cbfa03da51b-S1","host":"192.168.56.101","startedAt":"2016-10-15T21:21:06.234Z","stagedAt":"2016-10-15T21:20:55.709Z","ports":[31299],"version":"2016-10-15T21:20:55.434Z","ipAddresses":[{"ipAddress":"172.17.0.2","protocol":"IPv4"}],"appId":"/tomcat"},{"id":"tomcat.4263a55f-931d-11e6-b002-9e1a99d79ec2","slaveId":"81974b67-e403-47a0-bffd-7cbfa03da51b-S0","host":"192.168.56.103","startedAt":"2016-10-15T21:21:10.236Z","stagedAt":"2016-10-15T21:20:55.858Z","ports":[31204],"version":"2016-10-15T21:20:55.434Z","ipAddresses":[{"ipAddress":"172.17.0.2","protocol":"IPv4"}],"appId":"/tomcat"}]}}suser@ubuntu-1:~$

8. 总结

通过以上过程大家可以发现直接把docker容器起来就应用可以使用了，通过容器来运行服务简单而高效，省去了很多不必要的麻烦。相信随着越来越多的人使用docker技术，不断的推动docker技术的快速发展，docker技术会更加流行，应用会更加广泛。

tianpy5

关注

1
点赞
踩
3

收藏

觉得还不错? 一键收藏
0
评论
基于Mesos+Marathon+Zookeeper的Docker容器管理平台

刚接触docker的时候，发现docker很像一个虚拟机，随着使用的增多，它给我们开发工作带来了极大的方便，因此就想自己搭建一个docker集群环境，以便更深入了解docker集群的工作原理。之前曾经尝试分别安装mesos，marathon和zookeeper，组成分布式系统，但是在安装、配置过程中遇到了各种各样的问题，特别是mesos，问题比较多。既然docker容器的使用十分方便，所以我就使
复制链接

扫一扫

专栏目录