方式一: hadoop命令导出
hadoop fs -get hdfs://hadoop000:8020/data/page_views2 pv2
方式二:通过insert...directory导出 【spark暂不支持】
导出到本地：
INSERT OVERWRITE LOCAL directory ‘/home/spark/hivetmp/‘ ROW FORMAT DELIMITED FIELDS TERMINATED BY ‘\t‘ LINES TERMINATED BY ‘\n‘ select * from page_views;
导出到HDFS：
INSERT OVERWRITE directory ‘/hivetmp/‘ ROW FORMAT DELIMITED FIELDS TERMINATED BY ‘\t‘ LINES TERMINATED BY ‘\n‘ select * from page_views;
报错：cannot recognize input near ‘ROW‘ ‘FORMAT‘ ‘DELIMITED‘ in select clause
INSERT OVERWRITE directory ‘/hivetmp/‘ select * from page_views;
注意： 导出到本地可以通过ROW FORMAT来设置分隔符,导出到HDFS是不能设置分隔符的；
方式三: shell命令 + 管道(hive -f/-e | sed/grep/awk > file)
hive -e "select * from page_views limit 5" hive -S -e "select * from page_views limit 5" | grep B58W48U4WKZCJ5D1T3Z9ZY88RU7QA7B1 hive -S -e "select * from page_views limit 5" | grep B58W48U4WKZCJ5D1T3Z9ZY88RU7QA7B1 > file
方式四: sqoop
详见sqoop章节：http://www.cnblogs.com/luogankun/category/601761.html
原文：http://www.cnblogs.com/luogankun/p/4108482.html