spark history Server产生背景
以standalone运行模式为例，在运行Spark Application的时候，Spark会提供一个WEBUI列出应用程序的运行时信息；但该WEBUI随着Application的完成(成功/失败)而关闭，也就是说，Spark Application运行完(成功/失败)后，将无法查看Application的历史记录；
Spark history Server就是为了应对这种情况而产生的，通过配置可以在Application执行的过程中记录下了日志事件信息，那么在Application执行结束后，WEBUI就能重新渲染生成UI界面展现出该Application在执行过程中的运行时信息；
Spark运行在yarn或者mesos之上，通过spark的history server仍然可以重构出一个已经完成的Application的运行时参数信息（假如Application运行的事件日志信息已经记录下来）；
spark history Server的配置
1. 在Spark的conf目录下/usr/local/spark-1.6.0-bin-hadoop2.6/conf，将spark-defaults.conf.template改名为spark-defaults.conf
mv spark-defaults.conf.template spark-defaults.conf
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6$ ls
bin data examples licenses NOTICE README.md work
CHANGES.txt derby.log lib logs python RELEASE
conf ec2 LICENSE metastore_db R sbin
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6$ cd conf/
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6/conf$ ls
docker.properties.template metrics.properties.template spark-env.sh
fairscheduler.xml.template slaves
log4j.properties.template spark-defaults.conf.template
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6/conf$ mv spark-defaults.conf.template spark-defaults.conf
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6/conf$ ls
docker.properties.template metrics.properties.template spark-env.sh
fairscheduler.xml.template slaves
log4j.properties.template spark-defaults.conf
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6/conf$
2. 对spark-defaults.conf 配置
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6/conf$ vim spark-defaults.conf
spark.eventLog.enabled true
spark.eventLog.dir hdfs://SparkSingleNode:9000/historyserverforSpark
spark.history.ui.port 18080
spark.history.fs.logDirectory hdfs://SparkSingleNode:9000/historyserverforSpark
3.启动history-server
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6/conf$ cd ..
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6$ sbin/start-history-server.sh
starting org.apache.spark.deploy.history.HistoryServer, logging to /usr/local/spark/spark-1.5.2-bin-hadoop2.6/sbin/../logs/spark-spark-org.apache.spark.deploy.history.HistoryServer-1-SparkSingleNode.out
failed to launch org.apache.spark.deploy.history.HistoryServer:
full log in /usr/local/spark/spark-1.5.2-bin-hadoop2.6/sbin/../logs/spark-spark-org.apache.spark.deploy.history.HistoryServer-1-SparkSingleNode.out
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6$ jps
6816 Jps
3876 Worker
6772 HistoryServer
3174 NameNode
5990 CoarseGrainedExecutorBackend
3703 Master
3453 SecondaryNameNode
3293 DataNode
5887 SparkSubmit
spark@SparkSingleNode:/usr/local/spark/spark-1.5.2-bin-hadoop2.6$
4、spark-env.sh
5.在web浏览器中查看http://SparkSingleNode:18080/ 显示页面
1.6.0 History Server
Event log directory: hdfs://Master:9000/historyserverforSpark
成功！
本文转自大数据躺过的坑博客园博客，原文链接：http://www.cnblogs.com/zlslch/p/5922165.html，如需转载请自行联系原作者
