Spark集群配置-Mesosphere方案

#操作系统

  CentOS 7

  JDK 1.6以上， 如：/usr/share/jdk1.7.0_45

#修改节点名字

    本次搭建Spark集群机器主从节点， 修改host（/etc/hosts）如下：

	Hostname	Ip
Master	xd-ui	192.168.1.5
Slave 1	Xd-1	192.168.1.6
Slave 2	Xd-2	192.168.1.7
Slave 3	Xd-3	192.168.1.8

 

#安装mesosphere repo

sudo rpm -Uvh http://repos.mesosphere.io/el/7/noarch/RPMS/mesosphere-el-repo-7-1.noarch.rpm 

#下载apache mesos

wget http://downloads.mesosphere.io/master/centos/7/mesos-0.21.1-1.1.centos701406.x86_64.rpm

sudo rpm -Uvh mesos-0.21.1-1.1.centos701406.x86_64.rpm

#安装marathon

sudo yum -y install marathon

#安装chronos

sudo yum -y install chronos

 

ZooKeeper配置，此处略，详细看zook配置文档

On each node, replacing the IP addresses below with each master's IP address, set /etc/mesos/zk to:

zk://1.1.1.1:2181,2.2.2.2:2181,3.3.3.3:2181/mesos

#配置Mesosphere master节点

quorum

设置：/etc/mesos-master/quorum内容： 1

目前个人理解Quorum主节点个数，类似hadoop临时主节点，大家有意见的可以反馈我。

Hostname

设置/etc/mesos-master/hostname为xd-ui   

 

Work_dir

设置/etc/mesos-master/work_dir为工作目前，默认为：/var/lib/mesos，我这里设置/alidata1/mesos

 

配置结构显示如下

[dev@xd-ui ~]$ tree /etc/mesos-master/   /etc/mesos-master/   ├── hostname     à xd-ui   ├── quorum      à 1   └── work_dir    à /alidata1/mesos

重启Mesos Master:

[dev@xd-ui ~]$sudo service mesos-master   restart

#配置Mesosphere slave节点

Xd-1

配置各个节点的Slaves   [dev@spark-1 ~]$ tree /etc/mesos-slave/   /etc/mesos-slave/   ├── hostname    xd-1   └──   work_dir    /alidata1/mesos

Xd-2

配置各个节点的Slaves   [dev@spark-2 ~]$ tree /etc/mesos-slave/   /etc/mesos-slave/   ├── hostname    xd-2   └──   work_dir    /alidata1/mesos

Xd-3

配置各个节点的Slaves   [dev@spark-3 ~]$ tree /etc/mesos-slave/   /etc/mesos-slave/   ├── hostname    xd-3   └── work_dir    /alidata1/mesos

在节点[xd-1，xd-2，xd-3]上重启各个Mesos Slaves:

[dev@xd-1 ~]$sudo service   mesos-slave restart

 

检查各个节点是否起来

[dev@xd-ui ~]$sudo ps -ef | grep mesos /usr/sbin/mesos-slave   --master=zk://xd-1:2181,xd-2:2181,xd-3:2181/mesos --log_dir=/var/log/mesos   --hostname=xd-3 --work_dir=/alidata1/mesos

出现类似内容则启动正常

#访问Mesosphere主页,确认各个节点是否正常

浏览器输入 http://xd-ui:5050/，点击slaves标签页

#Spark配置

下载spark分布式二进制套件

注意：按实际需要，选择对应的spark和hadoop版本

解压spark-1.2.1-bin-hadoop2.4.tgz

tar -czvf   spark-1.2.1-bin-hadoop2.4.tgz spark-1.2.1-bin-hadoop2.4

修改${SPARK_HOME}/bin/spark-class或者conf/spark-env.sh，在首行添加：

export JAVA_HOME=/usr/share/jdk1.7.0_45

确定所有的Mesos 节点的JDK，按照目录是一致的.

再压缩成spark-1.2.1-bin-hadoop2.4.tgz

tar -czvf spark-1.2.1-bin-haoop2.4.tgz

 

再发布压缩tgz包到hdfs或者http。

可以使用Nginx做一个简单的下载服务器，比如我的下载地址是：

http://192.168.0.7/download/ spark-1.2.1-bin-haoop2.4.tgz

把解压的spark目录放到mesos-master上。

并且配置${SPARK_HOME}/conf/spark-env.sh，在文件末尾添加：

export MESOS_NATIVE_LIBRARY=/usr/local/lib/libmesos.so   export SPARK_EXECUTOR_URI=http://192.168.0.7/download/ spark-1.2.1-bin-haoop2.4.tgz   export MASTER=mesos://xd-ui:5050

 

#启动spark shell

 [dev@xd-ui spark-1.2.0-dist]$ ./bin/spark-shell

Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties

15/01/22 16:07:46 INFO SecurityManager: Changing view acls to: dev

15/01/22 16:07:46 INFO SecurityManager: Changing modify acls to: dev

15/01/22 16:07:46 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(dev); users with modify permissions: Set(dev)

15/01/22 16:07:46 INFO HttpServer: Starting HTTP Server

15/01/22 16:07:46 INFO Utils: Successfully started service 'HTTP class server' on port 50256.

Welcome to

      ____              __

     / __/__  ___ _____/ /__

    _\ \/ _ \/ _ `/ __/  '_/

   /___/ .__/\_,_/_/ /_/\_\   version 1.2.0

      /_/

 

Using Scala version 2.10.4 (Java HotSpot(TM) 64-Bit Server VM, Java 1.7.0_45)

Type in expressions to have them evaluated.

 

 

输入测试代码

scala> val data = 1 to 10000; val distData = sc.parallelize(data); distData.filter(_< 10).collect()

data: scala.collection.immutable.Range.Inclusive = Range(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170...

 

#查看mesosphere ui界面确定任务是否运行成功

浏览器输入http://xd-ui:5050

 

#打开spark WebUI界面查看任务执行结果：

浏览器输入：http://xd-ui:4040

 

参考资料

http://mesosphere.com/downloads/details/index.html http://mesosphere.com/docs/tutorials/run-spark-on-mesos/ http://mesosphere.com/docs/getting-started/datacenter/install/ http://mesos.apache.org/gettingstarted/ http://spark.apache.org/docs/latest/running-on-mesos.html#configuration