Hadoop 编程

MapReduce - 编程

处理

1. select：直接分析输入数据，取出需要的字段数据即可
2. where: 也是对输入数据处理的过程中进行处理，判断是否需要该数据
3. aggregation:min, max, sum
4. group by: 通过Reducer实现
5. sort
6. join: map join, reduce join

Third-Party Libraries

export LIBJARS=$MYLIB/commons-lang-2.3.jar, hadoop jar prohadoop-0.0.1-SNAPSHOT.jar org.aspress.prohadoop.c3. WordCountUsingToolRunner -libjars $LIBJARS

hadoop jar prohadoop-0.0.1-SNAPSHOT-jar-with-dependencies.jar org.aspress.prohadoop.c3. WordCountUsingToolRunner The dependent libraries are now included inside the application JAR file

一般还是上面的好，指定依赖可以利用Public Cache，如果是包含依赖，则每次都需要拷贝

参考书籍

MapReduce Design Patterns