Hadoop的安装和配置可以参考我之前的文章：在Win7虚拟机下搭建Hadoop2.6.0伪分布式环境。
本篇介绍如何在Hadoop2.6.0基础上搭建spark1.4.0单机环境。
1. 软件准备
scala-2.11.7.tgz
spark-1.4.0-bin-hadoop2.6.tgz
都可以从官网下载。
2. scala安装和配置
scala-2.11.7.tgz解压缩即可。我解压缩到目录/home/vm/tools/scala，之后配置~/.bash_profile环境变量。
|
#scala export SCALA_HOME=/home/vm/tools/scala export PATH=$SCALA_HOME/bin:$PATH |
使用source ~/.bash_profile生效。
验证scala安装是否成功：
交互式使用scala：
3. spark安装和配置
解压缩spark-1.4.0-bin-hadoop2.6.tgz到/home/vm/tools/spark目录，之后配置~/.bash_profile环境变量。
|
#spark export SPARK_HOME=/home/vm/tools/spark export PATH=$SPARK_HOME/bin:$PATH |
修改$SPARK_HOME/conf/spark-env.sh
|
export SPARK_HOME=/home/vm/tools/spark export SCALA_HOME=/home/vm/tools/scala export JAVA_HOME=/home/vm/tools/jdk export SPARK_MASTER_IP=192.168.62.129 export SPARK_WORKER_MEMORY=512m |
修改$SPARK_HOME/conf/spark-defaults.conf
|
spark.master spark://192.168.62.129:7077 spark.serializer org.apache.spark.serializer.KryoSerializer |
修改$SPARK_HOME/conf/spark-defaults.conf
|
192.168.62.129 这是我测试机器的IP地址 |
启动spark
|
cd /home/vm/tools/spark/sbin sh start-all.sh |
测试Spark是否安装成功
|
cd $SPARK_HOME/bin/ ./run-example SparkPi |
SparkPi的执行日志：
1 vm@ubuntu:~/tools/spark/bin$ ./run-example SparkPi 2 3 Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties 4 5 15/07/29 00:02:32 INFO SparkContext: Running Spark version 1.4.0 6 7 15/07/29 00:02:33 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable 8 9 15/07/29 00:02:34 INFO SecurityManager: Changing view acls to: vm 10 11 15/07/29 00:02:34 INFO SecurityManager: Changing modify acls to: vm 12 13 15/07/29 00:02:34 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(vm); users with modify permissions: Set(vm) 14 15 15/07/29 00:02:35 INFO Slf4jLogger: Slf4jLogger started 16 17 15/07/29 00:02:35 INFO Remoting: Starting remoting 18 19 15/07/29 00:02:36 INFO Remoting: Remoting started; listening on addresses :[akka.tcp://sparkDriver@192.168.62.129:34337] 20 21 15/07/29 00:02:36 INFO Utils: Successfully started service 'sparkDriver' on port 34337. 22 23 15/07/29 00:02:36 INFO SparkEnv: Registering MapOutputTracker 24 25 15/07/29 00:02:36 INFO SparkEnv: Registering BlockManagerMaster 26 27 15/07/29 00:02:36 INFO DiskBlockManager: Created local directory at /tmp/spark-78277899-e4c4-4dcc-8c16-f46fce5e657d/blockmgr-be03da6d-31fe-43dd-959c-6cfa4307b269 28 29 15/07/29 00:02:36 INFO MemoryStore: MemoryStore started with capacity 267.3 MB 30 31 15/07/29 00:02:36 INFO HttpFileServer: HTTP File server directory is /tmp/spark-78277899-e4c4-4dcc-8c16-f46fce5e657d/httpd-fdc26a4d-c0b6-4fc9-9dee-fb085191ee5a 32 33 15/07/29 00:02:36 INFO HttpServer: Starting HTTP Server 34 35 15/07/29 00:02:37 INFO Utils: Successfully started service 'HTTP file server' on port 56880. 36 37 15/07/29 00:02:37 INFO SparkEnv: Registering OutputCommitCoordinator 38 39 15/07/29 00:02:37 INFO Utils: Successfully started service 'SparkUI' on port 4040. 40 41 15/07/29 00:02:37 INFO SparkUI: Started SparkUI at http://192.168.62.129:4040 42 43 15/07/29 00:02:40 INFO SparkContext: Added JAR file:/home/vm/tools/spark/lib/spark-examples-1.4.0-hadoop2.6.0.jar at http://192.168.62.129:56880/jars/spark-examples-1.4.0-hadoop2.6.0.jar with timestamp 1438099360726 44 45 15/07/29 00:02:41 INFO Executor: Starting executor ID driver on host localhost 46 47 15/07/29 00:02:41 INFO Utils: Successfully started service 'org.apache.spark.network.netty.NettyBlockTransferService' on port 44722. 48 49 15/07/29 00:02:41 INFO NettyBlockTransferService: Server created on 44722 50 51 15/07/29 00:02:41 INFO BlockManagerMaster: Trying to register BlockManager 52 53 15/07/29 00:02:41 INFO BlockManagerMasterEndpoint: Registering block manager localhost:44722 with 267.3 MB RAM, BlockManagerId(driver, localhost, 44722) 54 55 15/07/29 00:02:41 INFO BlockManagerMaster: Registered BlockManager 56 57 15/07/29 00:02:43 INFO SparkContext: Starting job: reduce at SparkPi.scala:35 58 59 15/07/29 00:02:43 INFO DAGScheduler: Got job 0 (reduce at SparkPi.scala:35) with 2 output partitions (allowLocal=false) 60 61 15/07/29 00:02:43 INFO DAGScheduler: Final stage: ResultStage 0(reduce at SparkPi.scala:35) 62 63 15/07/29 00:02:43 INFO DAGScheduler: Parents of final stage: List() 64 65 15/07/29 00:02:43 INFO DAGScheduler: Missing parents: List() 66 67 15/07/29 00:02:43 INFO DAGScheduler: Submitting ResultStage 0 (MapPartitionsRDD[1] at map at SparkPi.scala:31), which has no missing parents 68 69 15/07/29 00:02:44 INFO MemoryStore: ensureFreeSpace(1888) called with curMem=0, maxMem=280248975 70 71 15/07/29 00:02:44 INFO MemoryStore: Block broadcast_0 stored as values in memory (estimated size 1888.0 B, free 267.3 MB) 72 73 15/07/29 00:02:44 INFO MemoryStore: ensureFreeSpace(1186) called with curMem=1888, maxMem=280248975 74 75 15/07/29 00:02:44 INFO MemoryStore: Block broadcast_0_piece0 stored as bytes in memory (estimated size 1186.0 B, free 267.3 MB) 76 77 15/07/29 00:02:44 INFO BlockManagerInfo: Added broadcast_0_piece0 in memory on localhost:44722 (size: 1186.0 B, free: 267.3 MB) 78 79 15/07/29 00:02:44 INFO SparkContext: Created broadcast 0 from broadcast at DAGScheduler.scala:874 80 81 15/07/29 00:02:44 INFO DAGScheduler: Submitting 2 missing tasks from ResultStage 0 (MapPartitionsRDD[1] at map at SparkPi.scala:31) 82 83 15/07/29 00:02:44 INFO TaskSchedulerImpl: Adding task set 0.0 with 2 tasks 84 85 15/07/29 00:02:45 INFO TaskSetManager: Starting task 0.0 in stage 0.0 (TID 0, localhost, PROCESS_LOCAL, 1344 bytes) 86 87 15/07/29 00:02:45 INFO TaskSetManager: Starting task 1.0 in stage 0.0 (TID 1, localhost, PROCESS_LOCAL, 1346 bytes) 88 89 15/07/29 00:02:45 INFO Executor: Running task 1.0 in stage 0.0 (TID 1) 90 91 15/07/29 00:02:45 INFO Executor: Running task 0.0 in stage 0.0 (TID 0) 92 93 15/07/29 00:02:45 INFO Executor: Fetching http://192.168.62.129:56880/jars/spark-examples-1.4.0-hadoop2.6.0.jar with timestamp 1438099360726 94 95 15/07/29 00:02:45 INFO Utils: Fetching http://192.168.62.129:56880/jars/spark-examples-1.4.0-hadoop2.6.0.jar to /tmp/spark-78277899-e4c4-4dcc-8c16-f46fce5e657d/userFiles-27c8dd76-e417-4d13-9bfd-a978cbbaacd1/fetchFileTemp5302506499464337647.tmp 96 97 15/07/29 00:02:47 INFO Executor: Adding file:/tmp/spark-78277899-e4c4-4dcc-8c16-f46fce5e657d/userFiles-27c8dd76-e417-4d13-9bfd-a978cbbaacd1/spark-examples-1.4.0-hadoop2.6.0.jar to class loader 98 99 15/07/29 00:02:47 INFO Executor: Finished task 1.0 in stage 0.0 (TID 1). 582 bytes result sent to driver 100 101 15/07/29 00:02:47 INFO Executor: Finished task 0.0 in stage 0.0 (TID 0). 582 bytes result sent to driver 102 103 15/07/29 00:02:47 INFO TaskSetManager: Finished task 1.0 in stage 0.0 (TID 1) in 2641 ms on localhost (1/2) 104 105 15/07/29 00:02:47 INFO TaskSetManager: Finished task 0.0 in stage 0.0 (TID 0) in 2718 ms on localhost (2/2) 106 107 15/07/29 00:02:47 INFO DAGScheduler: ResultStage 0 (reduce at SparkPi.scala:35) finished in 2.817 s 108 109 15/07/29 00:02:47 INFO TaskSchedulerImpl: Removed TaskSet 0.0, whose tasks have all completed, from pool 110 111 15/07/29 00:02:47 INFO DAGScheduler: Job 0 finished: reduce at SparkPi.scala:35, took 4.244145 s 112 113 Pi is roughly 3.14622 114 115 15/07/29 00:02:47 INFO SparkUI: Stopped Spark web UI at http://192.168.62.129:4040 116 117 15/07/29 00:02:47 INFO DAGScheduler: Stopping DAGScheduler 118 119 15/07/29 00:02:48 INFO MapOutputTrackerMasterEndpoint: MapOutputTrackerMasterEndpoint stopped! 120 121 15/07/29 00:02:48 INFO Utils: path = /tmp/spark-78277899-e4c4-4dcc-8c16-f46fce5e657d/blockmgr-be03da6d-31fe-43dd-959c-6cfa4307b269, already present as root for deletion. 122 123 15/07/29 00:02:48 INFO MemoryStore: MemoryStore cleared 124 125 15/07/29 00:02:48 INFO BlockManager: BlockManager stopped 126 127 15/07/29 00:02:48 INFO BlockManagerMaster: BlockManagerMaster stopped 128 129 15/07/29 00:02:48 INFO OutputCommitCoordinator$OutputCommitCoordinatorEndpoint: OutputCommitCoordinator stopped! 130 131 15/07/29 00:02:48 INFO SparkContext: Successfully stopped SparkContext 132 133 15/07/29 00:02:48 INFO Utils: Shutdown hook called 134 135 15/07/29 00:02:48 INFO Utils: Deleting directory /tmp/spark-78277899-e4c4-4dcc-8c16-f46fce5e657d
在浏览器中打开地址 http://192.168.62.129:8080 可以查看spark集群和任务基本情况：
4. spark-shell工具
在/home/vm/tools/spark/bin下执行./spark-shell，即可进入spark-shell交互界面。通过spark-shell可以进行一些调试工作。
1 vm@ubuntu:~/tools/spark/bin$ ./spark-shell 2 3 log4j:WARN No appenders could be found for logger (org.apache.hadoop.metrics2.lib.MutableMetricsFactory). 4 5 log4j:WARN Please initialize the log4j system properly. 6 7 log4j:WARN See http://logging.apache.org/log4j/1.2/faq.html#noconfig for more info. 8 9 Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties 10 11 15/07/29 00:08:02 INFO SecurityManager: Changing view acls to: vm 12 13 15/07/29 00:08:02 INFO SecurityManager: Changing modify acls to: vm 14 15 15/07/29 00:08:02 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(vm); users with modify permissions: Set(vm) 16 17 15/07/29 00:08:03 INFO HttpServer: Starting HTTP Server 18 19 15/07/29 00:08:04 INFO Utils: Successfully started service 'HTTP class server' on port 56464. 20 21 Welcome to 22 23 ____ __ 24 25 / __/__ ___ _____/ /__ 26 27 _ / _ / _ `/ __/ '_/ 28 29 /___/ .__/\_,_/_/ /_/\_ version 1.4.0 30 31 /_/ 32 33 34 35 Using Scala version 2.10.4 (Java HotSpot(TM) 64-Bit Server VM, Java 1.7.0_80) 36 37 Type in expressions to have them evaluated. 38 39 Type :help for more information. 40 41 15/07/29 00:08:37 INFO SparkContext: Running Spark version 1.4.0 42 43 15/07/29 00:08:38 INFO SecurityManager: Changing view acls to: vm 44 45 15/07/29 00:08:38 INFO SecurityManager: Changing modify acls to: vm 46 47 15/07/29 00:08:38 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(vm); users with modify permissions: Set(vm) 48 49 15/07/29 00:08:40 INFO Slf4jLogger: Slf4jLogger started 50 51 15/07/29 00:08:41 INFO Remoting: Starting remoting 52 53 15/07/29 00:08:42 INFO Remoting: Remoting started; listening on addresses :[akka.tcp://sparkDriver@192.168.62.129:59312] 54 55 15/07/29 00:08:42 INFO Utils: Successfully started service 'sparkDriver' on port 59312. 56 57 15/07/29 00:08:42 INFO SparkEnv: Registering MapOutputTracker 58 59 15/07/29 00:08:42 INFO SparkEnv: Registering BlockManagerMaster 60 61 15/07/29 00:08:43 INFO DiskBlockManager: Created local directory at /tmp/spark-621ebed4-8bd8-4e87-9ea5-08b5c7f05e98/blockmgr-a12211dd-e0ba-4556-999c-6249b9c44d8a 62 63 15/07/29 00:08:43 INFO MemoryStore: MemoryStore started with capacity 267.3 MB 64 65 15/07/29 00:08:43 INFO HttpFileServer: HTTP File server directory is /tmp/spark-621ebed4-8bd8-4e87-9ea5-08b5c7f05e98/httpd-8512d909-5a81-4935-8fbd-2b2ed741ae26 66 67 15/07/29 00:08:43 INFO HttpServer: Starting HTTP Server 68 69 15/07/29 00:08:57 INFO Utils: Successfully started service 'HTTP file server' on port 43678. 70 71 15/07/29 00:09:00 INFO SparkEnv: Registering OutputCommitCoordinator 72 73 15/07/29 00:09:02 INFO Utils: Successfully started service 'SparkUI' on port 4040. 74 75 15/07/29 00:09:02 INFO SparkUI: Started SparkUI at http://192.168.62.129:4040 76 77 15/07/29 00:09:03 INFO Executor: Starting executor ID driver on host localhost 78 79 15/07/29 00:09:03 INFO Executor: Using REPL class URI: http://192.168.62.129:56464 80 81 15/07/29 00:09:04 INFO Utils: Successfully started service 'org.apache.spark.network.netty.NettyBlockTransferService' on port 50636. 82 83 15/07/29 00:09:04 INFO NettyBlockTransferService: Server created on 50636 84 85 15/07/29 00:09:04 INFO BlockManagerMaster: Trying to register BlockManager 86 87 15/07/29 00:09:04 INFO BlockManagerMasterEndpoint: Registering block manager localhost:50636 with 267.3 MB RAM, BlockManagerId(driver, localhost, 50636) 88 89 15/07/29 00:09:04 INFO BlockManagerMaster: Registered BlockManager 90 91 15/07/29 00:09:05 INFO SparkILoop: Created spark context.. 92 93 Spark context available as sc. 94 95 15/07/29 00:09:07 INFO HiveContext: Initializing execution hive, version 0.13.1 96 97 15/07/29 00:09:09 INFO HiveMetaStore: 0: Opening raw store with implemenation class:org.apache.hadoop.hive.metastore.ObjectStore 98 99 15/07/29 00:09:09 INFO ObjectStore: ObjectStore, initialize called 100 101 15/07/29 00:09:10 INFO Persistence: Property datanucleus.cache.level2 unknown - will be ignored 102 103 15/07/29 00:09:10 INFO Persistence: Property hive.metastore.integral.jdo.pushdown unknown - will be ignored 104 105 15/07/29 00:09:11 WARN Connection: BoneCP specified but not present in CLASSPATH (or one of dependencies) 106 107 15/07/29 00:09:13 WARN Connection: BoneCP specified but not present in CLASSPATH (or one of dependencies) 108 109 15/07/29 00:09:19 INFO ObjectStore: Setting MetaStore object pin classes with hive.metastore.cache.pinobjtypes="Table,StorageDescriptor,SerDeInfo,Partition,Database,Type,FieldSchema,Order" 110 111 15/07/29 00:09:20 INFO MetaStoreDirectSql: MySQL check failed, assuming we are not on mysql: Lexical error at line 1, column 5. Encountered: "@" (64), after : "". 112 113 15/07/29 00:09:23 INFO Datastore: The class "org.apache.hadoop.hive.metastore.model.MFieldSchema" is tagged as "embedded-only" so does not have its own datastore table. 114 115 15/07/29 00:09:23 INFO Datastore: The class "org.apache.hadoop.hive.metastore.model.MOrder" is tagged as "embedded-only" so does not have its own datastore table. 116 117 15/07/29 00:09:28 INFO Datastore: The class "org.apache.hadoop.hive.metastore.model.MFieldSchema" is tagged as "embedded-only" so does not have its own datastore table. 118 119 15/07/29 00:09:28 INFO Datastore: The class "org.apache.hadoop.hive.metastore.model.MOrder" is tagged as "embedded-only" so does not have its own datastore table. 120 121 15/07/29 00:09:29 INFO ObjectStore: Initialized ObjectStore 122 123 15/07/29 00:09:29 WARN ObjectStore: Version information not found in metastore. hive.metastore.schema.verification is not enabled so recording the schema version 0.13.1aa 124 125 15/07/29 00:09:31 INFO HiveMetaStore: Added admin role in metastore 126 127 15/07/29 00:09:31 INFO HiveMetaStore: Added public role in metastore 128 129 15/07/29 00:09:31 INFO HiveMetaStore: No user is added in admin role, since config is empty 130 131 15/07/29 00:09:32 INFO SessionState: No Tez session required at this point. hive.execution.engine=mr. 132 133 15/07/29 00:09:32 INFO SparkILoop: Created sql context (with Hive support).. 134 135 SQL context available as sqlContext. 136 137 138 139 scala>
下一篇将介绍分别用eclipse和IDEA搭建spark开发环境。