impala的入门使用

1 使用

1.1

• 准备数据

  user.csv

  392456197008193000,张三,20,0
  267456198006210000,李四,25,1
  892456199007203000,王五,24,1
  492456198712198000,赵六,26,2
  392456197008193000,张三,20,0
  392456197008193000,张三,20,0
  

• 上传数据

  hadoop fs -mkdir -p /data/impala/t1
  #上传本地user.csv到hdfs /user/impala/table1
  hadoop fs -put user.csv /user/impala/t1
  

• 创建表

  #进⼊impala-shell
  impala-shell
  #表如果存在则删除
  drop table if exists t1;
  #执⾏创建
  create external table t1(id string,name string,age int,gender int) row format delimited fields terminated by ',' location '/data/impala/t1';
  

• 查询数据

  [lew3:21000] > select * from t1;
  Query: select * from t1
  +--------------------+------+-----+--------+
  | id                 | name | age | gender |
  +--------------------+------+-----+--------+
  | 392456197008193000 | 张三 | 20  | 0      |
  | 267456198006210000 | 李四 | 25  | 1      |
  | 892456199007203000 | 王五 | 24  | 1      |
  | 492456198712198000 | 赵六 | 26  | 2      |
  | 392456197008193000 | 张三 | 20  | 0      |
  | 392456197008193000 | 张三 | 20  | 0      |
  +--------------------+------+-----+--------+
  

• 创建t2表

  #创建⼀个内部表
  create table t2(id string,name string,age int,gender int) row format delimited fields terminated by ',';
  #查看表结构
  desc t1;
  desc formatted t2;
  

• 插⼊数据到t2

  insert overwrite table t2 select * from t1 where gender =0; #验证数据
  [lew3:21000] > select * from t2;
  Query: select * from t2
  +--------------------+------+-----+--------+
  | id                 | name | age | gender |
  +--------------------+------+-----+--------+
  | 392456197008193000 | 张三 | 20  | 0      |
  | 392456197008193000 | 张三 | 20  | 0      |
  | 392456197008193000 | 张三 | 20  | 0      |
  +--------------------+------+-----+--------+