impala

impala

1.Import csv file to hdfs:
export HADOOP_USER_NAME=impala
./hadoop fs -put /Users/gennadii/dev/performance_impala.csv hdfs://<host>:<port>/user/impala/performance_impala.csv

2.Create IDEA datasource Impala using Impala JDBC driver (all jars in zip)

3.Create connection
jdbc:impala://<host>:<port>

4.Create table (the same name as file):
create table performance_impala(
  field_string_1 string,
  field_integer_1 int,
  field_double_1 double,
  field_date_1 timestamp)
  row format delimited
  fields terminated by ',';
  
5.Load data from csv (use absolute hdfs path)
LOAD DATA INPATH '/user/impala/performance_impala.csv' INTO TABLE performance_impala

6.Check:
select * from performance_impala limit 10;

7.Skip 1st header row:
alter table performance_impala set tblproperties('skip.header.line.count'='1');