hive techooloogy log

1. download: http://mirror.esocc.com/apache/hive/stable/hive-0.11.0-bin.tar.gz

2.  use winscp upload to centos 6  & tar xzvf hive-*.tar.gz

3. ${HIVE_HOME}/bin/hive

        set hive.exec.scratchdir=/tmp/hive;

        set fs.default.name=hdfs://ip:9000;

        set hadoop.config.dir=/opt/hadoop-2.2.0/etc/hadoop;

       //合并小文件, 减小map

       set mapred.max.split.size=100000000;
       set mapred.min.split.size.per.node=100000000;
       set mapred.min.split.size.per.rack=100000000;
       set hive.input.format=org.apache.hadoop.hive.ql.io.CombineHiveInputFormat;

      //文件大、行数多，可考虑增加map数量

      set mapred.reduce.tasks=10;

    

4.  hive>

           create table hivetest (time BIGINT, sitename String, devname string, type string, content string) row format delimited fields terminated by ',';

           select count (*)  from hivetest;

          create table hivetest_1 as select * from hivetest distribute by rand(123);

          drop table hivetest_1;