WeeHive

A minimal-as-possible Docker container running Apache Hive on Hadoop. Intended for non-production use cases like testing out Hive code or running integration tests.

Setup

1. Install Docker.

2. Make sure that you have at least a few GB of memory allocated to Docker. Instructions:

   • Docker for Mac
   • Docker for Windows

3. Clone this repository.

4. From the repository root, build the Docker image.

   docker build -t weehive .

Usage

Beeline

docker run --rm -it \
  -v weehive_hadoop:/usr/local/hadoop/warehouse \
  -v weehive_meta:/usr/local/hadoop/metastore_db \
  weehive

You will be shown the Beeline shell. The weehive_hadoop and weehive_meta volume names can be changed to be project-specific names if you want.

Remote connection

1. Run the server.

   docker run --rm -it -p 10000:10000 \
     -v weehive_hadoop:/usr/local/hadoop/warehouse \
     -v weehive_meta:/usr/local/hadoop/metastore_db \
     weehive hiveserver2

2. Wait ~90 seconds for Hive to fully start.

3. Connect using the JDBC URL jdbc:hive2://localhost:10000. Example from an external beeline:

   beeline -u jdbc:hive2://localhost:10000

Loading data from file

1. Mount the data as a volume by adding a -v <sourcedir>:/usr/local/hadoop/data to one of the docker run commands above.
2. Follow instructions to load data

Development

docker build -t weehive:local .
docker run --rm -it weehive:local