Spark运行环境的安装(Standalone)

Spark功能还是蛮强的，安装的东西可是不少，好在搞完一次就可以一直用（除非用不上）。这里介绍安装需要的软件和步骤。不同机器可能还有些设置不一样的，需要自己去摸索，毕竟这个是开源软件，好事是有问题可以看源代码，坏事也是有问题了要去看源代码。

1、准备工作

    scala-2.9.3：一种编程语言，下载地址：http://www.scala-lang.org/download/

    spark-1.4.0：必须是编译好的Spark，如果下载的是Source，则需要自己根据环境使用SBT或者MAVEN重新编译才能使用。  

    编译好的 Spark下载地址：http://spark.apache.org/downloads.html。

2、安装scala-2.9.3

#解压scala-2.9.3.tgz
tar -zxvf scala-2.9.3.tgz
#配置SCALA_HOME
vi /etc/profile
#添加如下环境
export SCALA_HOME=/home/apps/scala-2.9.3
export PATH=.:$SCALA_HOME/bin:$PATH
#测试scala安装是否成功
#直接输入
scala

3、安装spark-1.4.0

#解压spark-1.4.0.tgz
tar -zxvf spark-1.4.0.tgz
#配置SPARK_HOME
vi /etc/profile
#添加如下环境
export SCALA_HOME=/home/apps/spark-1.4.0
export PATH=.:$SPARK_HOME/bin:$SPARK_HOME/sbin:$PATH

4、修改Spark配置文件

#复制slaves.template和 spark-env.sh.template各一份
cp  spark-env.sh.template  spark-env.sh
cp  slaves.template slaves
#slaves，此文件是指定子节点的主机,直接添加子节点主机名即可

    在spark-env.sh末端添加如下几行：

#JDK安装路径
export JAVA_HOME=/root/app/jdk
#SCALA安装路径
export SCALA_HOME=/root/app/scala-2.9.3
#主节点的IP地址
export SPARK_MASTER_IP=192.168.1.200
#分配的内存大小
export SPARK_WORKER_MEMORY=200m
#指定hadoop的配置文件目录
export HADOOP_CONF_DIR=/root/app/hadoop/etc/hadoop
#指定worker工作时分配cpu数量
export SPARK_WORKER_CORES=1
#指定spark实例，一般1个足以
export SPARK_WORKER_INSTANCES=1
#jvm操作，在spark1.0之后增加了spark-defaults.conf默认配置文件，该配置参数在默认配置在该文件中
export SPARK_JAVA_OPTS

    spark-defaults.conf中还有如下配置参数：

SPARK.MASTER //spark://hostname:8080
SPARK.LOCAL.DIR //spark工作目录（做shuffle的目录）
SPARK.EXECUTOR.MEMORY //spark1.0抛弃SPARK_MEM参数,使用该参数

5、测试spark安装是否成功

在主节点机器上启动顺序
1、先启动hdfs(./sbin/start-dfs.sh)
2、启动spark-master(./sbin/start-master.sh)
3、启动spark-worker(./sbin/start-slaves.sh)
4、jps查看进程有
主节点：namenode、secondrynamnode、master
从节点：datanode、worker
5、启动spark-shell
15/06/21 21:23:47 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
15/06/21 21:23:47 INFO spark.SecurityManager: Changing view acls to: root
15/06/21 21:23:47 INFO spark.SecurityManager: Changing modify acls to: root
15/06/21 21:23:47 INFO spark.SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(root); users with modify permissions: Set(root)
15/06/21 21:23:47 INFO spark.HttpServer: Starting HTTP Server
15/06/21 21:23:47 INFO server.Server: jetty-8.y.z-SNAPSHOT
15/06/21 21:23:47 INFO server.AbstractConnector: Started SocketConnector@0 .0.0.0:38651
15/06/21 21:23:47 INFO util.Utils: Successfully started service 'HTTP class server' on port 38651.
Welcome to
____              __
/ __/__  ___ _____/ /__
_\ \/ _ \/ _ `/ __/  '_/
/___/ .__/\_,_/_/ /_/\_\   version 1.4.0
/_/

Using Scala version 2.10.4 (Java HotSpot(TM) Client VM, Java 1.7.0_65)
Type in expressions to have them evaluated.
Type :help for more information.
15/06/21 21:23:54 INFO spark.SparkContext: Running Spark version 1.4.0
15/06/21 21:23:54 INFO spark.SecurityManager: Changing view acls to: root
15/06/21 21:23:54 INFO spark.SecurityManager: Changing modify acls to: root
15/06/21 21:23:54 INFO spark.SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(root); users with modify permissions: Set(root)
15/06/21 21:23:56 INFO slf4j.Slf4jLogger: Slf4jLogger started
15/06/21 21:23:56 INFO Remoting: Starting remoting
15/06/21 21:23:57 INFO Remoting: Remoting started; listening on addresses :[akka.tcp://sparkDriver@192.168.1.200:57658]
15/06/21 21:23:57 INFO util.Utils: Successfully started service 'sparkDriver' on port 57658.
15/06/21 21:23:58 INFO spark.SparkEnv: Registering MapOutputTracker
15/06/21 21:23:58 INFO spark.SparkEnv: Registering BlockManagerMaster
15/06/21 21:23:58 INFO storage.DiskBlockManager: Created local directory at /tmp/spark-4f1badf6-1e92-47ca-98a2-6d82f4882f15/blockmgr-530e4335-9e59-45d4-b9fb-6014089f5a00
15/06/21 21:23:58 INFO storage.MemoryStore: MemoryStore started with capacity 267.3 MB
15/06/21 21:23:59 INFO spark.HttpFileServer: HTTP File server directory is /tmp/spark-4f1badf6-1e92-47ca-98a2-6d82f4882f15/httpd-4b2cca3c-e8d4-4ab3-9c3d-38ec579ec873
15/06/21 21:23:59 INFO spark.HttpServer: Starting HTTP Server
15/06/21 21:23:59 INFO server.Server: jetty-8.y.z-SNAPSHOT
15/06/21 21:23:59 INFO server.AbstractConnector: Started SocketConnector@0 .0.0.0:51899
15/06/21 21:23:59 INFO util.Utils: Successfully started service 'HTTP file server' on port 51899.
15/06/21 21:23:59 INFO spark.SparkEnv: Registering OutputCommitCoordinator
15/06/21 21:23:59 INFO server.Server: jetty-8.y.z-SNAPSHOT
15/06/21 21:23:59 INFO server.AbstractConnector: Started SelectChannelConnector@0 .0.0.0:4040
15/06/21 21:23:59 INFO util.Utils: Successfully started service 'SparkUI' on port 4040.
15/06/21 21:23:59 INFO ui.SparkUI: Started SparkUI at http://192.168.1.200:4040 15/06/21 21:24:00 INFO executor.Executor: Starting executor ID driver on host localhost
15/06/21 21:24:00 INFO executor.Executor: Using REPL class URI: http://192.168.1.200:38651 15/06/21 21:24:01 INFO util.Utils: Successfully started service 'org.apache.spark.network.netty.NettyBlockTransferService' on port 59385.
15/06/21 21:24:01 INFO netty.NettyBlockTransferService: Server created on 59385
15/06/21 21:24:01 INFO storage.BlockManagerMaster: Trying to register BlockManager
15/06/21 21:24:01 INFO storage.BlockManagerMasterEndpoint: Registering block manager localhost:59385 with 267.3 MB RAM, BlockManagerId(driver, localhost, 59385)
15/06/21 21:24:01 INFO storage.BlockManagerMaster: Registered BlockManager
15/06/21 21:24:02 INFO repl.SparkILoop: Created spark context..
Spark context available as sc.
15/06/21 21:24:03 INFO hive.HiveContext: Initializing execution hive, version 0.13.1
15/06/21 21:24:04 INFO metastore.HiveMetaStore: 0: Opening raw store with implemenation class:org.apache.hadoop.hive.metastore.ObjectStore
15/06/21 21:24:04 INFO metastore.ObjectStore: ObjectStore, initialize called
15/06/21 21:24:04 INFO DataNucleus.Persistence: Property datanucleus.cache.level2 unknown - will be ignored
15/06/21 21:24:04 INFO DataNucleus.Persistence: Property hive.metastore.integral.jdo.pushdown unknown - will be ignored
15/06/21 21:24:05 WARN DataNucleus.Connection: BoneCP specified but not present in CLASSPATH (or one of dependencies)
15/06/21 21:24:07 WARN DataNucleus.Connection: BoneCP specified but not present in CLASSPATH (or one of dependencies)
15/06/21 21:24:14 INFO metastore.ObjectStore: Setting MetaStore object pin classes with hive.metastore.cache.pinobjtypes="Table,StorageDescriptor,SerDeInfo,Partition,Database,Type,FieldSchema,Order"
15/06/21 21:24:14 INFO metastore.MetaStoreDirectSql: MySQL check failed, assuming we are not on mysql: Lexical error at line 1, column 5.  Encountered: "@" (64), after : "".
15/06/21 21:24:15 INFO DataNucleus.Datastore: The class "org.apache.hadoop.hive.metastore.model.MFieldSchema" is tagged as "embedded-only" so does not have its own datastore table.
15/06/21 21:24:15 INFO DataNucleus.Datastore: The class "org.apache.hadoop.hive.metastore.model.MOrder" is tagged as "embedded-only" so does not have its own datastore table.
15/06/21 21:24:18 INFO DataNucleus.Datastore: The class "org.apache.hadoop.hive.metastore.model.MFieldSchema" is tagged as "embedded-only" so does not have its own datastore table.
15/06/21 21:24:18 INFO DataNucleus.Datastore: The class "org.apache.hadoop.hive.metastore.model.MOrder" is tagged as "embedded-only" so does not have its own datastore table.
15/06/21 21:24:19 INFO metastore.ObjectStore: Initialized ObjectStore
15/06/21 21:24:20 WARN metastore.ObjectStore: Version information not found in metastore. hive.metastore.schema.verification is not enabled so recording the schema version 0.13.1aa
15/06/21 21:24:24 INFO metastore.HiveMetaStore: Added admin role in metastore
15/06/21 21:24:24 INFO metastore.HiveMetaStore: Added public role in metastore
15/06/21 21:24:24 INFO metastore.HiveMetaStore: No user is added in admin role, since config is empty
15/06/21 21:24:25 INFO session.SessionState: No Tez session required at this point. hive.execution.engine=mr.
15/06/21 21:24:25 INFO repl.SparkILoop: Created sql context (with Hive support)..
SQL context available as sqlContext.

6、使用wordcount例子测试，启动spark-shell之前先上传一份文件到hdfs
7、代码：
val file = sc.textFile("hdfs://hadoop.master:9000/data/intput/wordcount.data")
val count = file.flatMap(line=>(line.split(" "))).map(word=>(word,1)).reduceByKey(_+_)
count.collect()
count.textAsFile("hdfs://hadoop.master:9000/data/output")
理解上面的代码你需要学习scala语言。

直接打印结果：hadoop dfs -cat /data/output/p*
(im,1)
(are,1)
(yes,1)
(hi,2)
(do,1)
(no,3)
(to,1)
(lll,1)
(,3)
(hello,3)
(xiaoming,1)
(ga,1)
(world,1)