Categories

  • Big Data

Tags

  • hbase
  • hive

Hive Query Over Hbase

Hive and HBase are two powerful technologies that, when combined, can provide a comprehensive solution for big data storage and analysis. In this blog post, we’ll walk you through the process of executing Hive queries over HBase. This integration enables you to leverage Hive’s SQL-like querying capabilities on top of HBase’s NoSQL database. Let’s get started with the step-by-step guide.

  1. Launch the hbase shell-
hbase shell
  1. Create employee table with column family ‘cf’-

create ‘employees’, ‘cf’

  1. Let’s populate the “employees” table with some sample data. Use the following commands to add employee details:
put 'employees','1001','cf:name','mike'

put 'employees','1001','cf:country','USA'

put 'employees','1002','cf:name','sky'

put 'employees','1002','cf:country','Canada'

put 'employees','1003','cf:name','sara'

put 'employees','1003','cf:country','aus'

  1. Launch the Hive CLI and Define an External Table Now switch to the Hive environment. Launch the Hive CLI using the hive command. Once you’re in the Hive CLI, you can define an external table that references the HBase table. Enter the following HiveQL query:
CREATE EXTERNAL TABLE hbase_employee(key STRING, name STRING, country STRING)
STORED BY 'org.apache.hadoop.hive.hbase.HBaseStorageHandler'
WITH SERDEPROPERTIES ("hbase.columns.mapping" = ":key,cf:name,cf:country")
TBLPROPERTIES ("hbase.table.name" = "employees");

This HiveQL query defines an external table named “hbase_employee” that points to the “employees” HBase table. It specifies the column mapping and other properties required for integrating Hive with HBase.

  1. Test Your Hive Query

With the external table defined, you can now run Hive queries on HBase data. For example, you can run a simple query to count employees by country:

select country,count(\*) from hbase\_employee group by country;

Tip: Handling HBase Classes Not Found Exception If you encounter an HBase classes not found exception while running Hive queries, it may be due to missing HBase JAR files. To resolve this issue, you can add the HBase JAR to Hive’s auxiliary path. Here’s the command to do so:

hive --auxpath ${HBASE_HOME}/hbase-0.94.13.jar

By following the steps outlined, you can seamlessly execute Hive queries over HBase, enabling you to work with both structured and unstructured data within your big data ecosystem. Cheers!