hive-metastore

Apache Hive Metastore as a Standalone server in Docker. Can be used in a modular fashion with Presto, Trino, Spark, and many other BigData tools.

There are numerous Docker images that attempt to do this, but yet to see something actually work as advertised with minimal bloat.

Setup

Usage

The recommended way to get this Docker Image is to pull the prebuilt image from the Docker Hub Registry.

$ docker pull naushadh/hive-metastore

If you wish, you can also build the image yourself by cloning the repository, and executing the docker build command.

$ git clone https://github.com/naushadh/hive-metastore
$ cd hive-metastore
$ make build

You can also use DockerCompose to launch the app with all of its dependencies using docker-compose.yml file in the GitHub repository. Just replace the build with image

app:
-   build: .
+   image: naushadh/hive-metastore

And then run

$ docker compose up -d
[+] Running 4/4
 ✔ Container hive-metastore-localstack-1  Healthy                                                                   10.8s 
 ✔ Container hive-metastore-postgres-1    Healthy                                                                   30.8s 
 ✔ Container hive-metastore-s3_setup-1    Started                                                                   11.0s 
 ✔ Container hive-metastore-app-1         Started                                                                   31.0s 
$ docker compose ps
NAME                          IMAGE                       COMMAND                  SERVICE             CREATED              STATUS                    PORTS
hive-metastore-app-1          hive-metastore-app          "./run.sh"               app                 About a minute ago   Up 8 seconds              0.0.0.0:9083->9083/tcp
hive-metastore-localstack-1   localstack/localstack:1.1   "docker-entrypoint.sh"   localstack          About a minute ago   Up 39 seconds (healthy)   4510-4559/tcp, 5678/tcp, 0.0.0.0:4566->4566/tcp
hive-metastore-postgres-1     postgres:14-alpine          "docker-entrypoint.s…"   postgres            About a minute ago   Up 39 seconds (healthy)   0.0.0.0:5432->5432/tcp

You can now connect to the MetaStore Thrift server at 0.0.0.0:9083 from your host machine.

Configuration

Controlled via ENVironment variables

Key	Required?	Description
DATABASE_TYPE_JDBC	No, defaults to postgresql	Database type1 for JDBC connection
DATABASE_TYPE	No, defaults to postgres	Database type1 for migration tool
DATABASE_DRIVER	No, defaults to org.postgresql.Driver	Database class used for JDBC connection
DATABASE_HOST	Yes	Database host
DATABASE_PORT	No, defaults to 5432	Database port
DATABASE_DB	Yes	Database name
DATABASE_USER	Yes	Database user
DATABASE_PASSWORD	Yes	Database password
S3_ENDPOINT_URL	No	Custom S3 endpoint URL; useful for LocalStack integration
S3_BUCKET	Yes	S3 bucket name
S3_PREFIX	Yes	S3 bucket prefix

¹ Though you have the ability to modify DATABASE_TYPE_JDBC/DATABASE_TYPE, we presently only install Postgres driver. You'd have to extend this image and install a non-Postgres driver to change the Database type.

Development

This project has most of the batteries included to test and verify that the app works

1. Install docker (27+) with compose

2. Launch dev environment

   $ make build env-up

3. Run test(s)

   $ make test