Running Hive Queries in Zeppelin

This section contains samples of Apache Hive queries that you can run in your Apache Zeppelin notebook.

Prerequisites

Before running Hive queries, make sure you have configured the Hive JDBC interpreter. Also, see MapR Data Science Refinery Support by MapR Core Version for limitations when connecting to a secure MapR Data Platform 6.1 cluster.

Procedure

  1. Using the shell interpreter, create a source data file: 
    %sh
cat > /tmp/test.data << EOF
John,Smith
Brian,May
Rodger,Taylor
John,Deacon
Max,Plank
Freddie,Mercury
Albert,Einstein
Fedor,Dostoevsky
Lev,Tolstoy
Niccolo,Paganini
EOF
  2. Copy the file to the MapR File System:
    To use POSIX shell commands like cp, you must have a MapR filesystem mount point in your container. The example below assumes your mount point is /mapr and your cluster name is my.cluster.com:
    %sh
cp /tmp/test.data /mapr/my.cluster.com/user/mapruser1
    %sh
hadoop fs -put /tmp/test.data /user/mapruser1
  3. Run the Hive code using the Hive JDBC interpreter: 
    %hive
-- create and load Hive table
create table test_hive(first_name string, last_name string) ROW FORMAT DELIMITED FIELDS TERMINATED BY ',';
load data inpath '/user/mapruser1/test.data' overwrite into table test_hive;
-- create and load Hive ORC table
create table test_hive_orc(first_name string, last_name string) stored as orc tblproperties ("orc.compress"="NONE");
insert overwrite table test_hive_orc  select * from test_hive;
-- query the Hive ORC table
select * from test_hive_orc;

    The output looks like the following:

  4. Drop the Hive tables created in the example: 
    %hive
drop table test_hive;
drop table test_hive_orc;