Setup

You can download Cottontail DB releases from the releases page or build the distributions from the source.

Using Cottontail DB Docker Container

There are pre-built Docker containers of Cottontail DB for the linux/amd64 (starting from version 0.12.5) and linux/arm64 (starting from version 0.13.3) platforms on DockerHub for every release version. You can run it using the following command

docker run -it -p 1865:1865 -v /path/to/folder/on/host:/cottontaildb-data -v /path/to/host/config.json:/cottontaildb-config/config.json vitrivr/cottontaildb:latest

Important #1: You should expose the Cottontail DB port using -p 1865:1865 (adjust if you're using a different port) and map the data directory from the host machine into the container using -v. If you do not map your own config.json into the container, the default config will be used.

Important #2: If you're running the container in detached mode -d, then you should switch-off the CLI in the config.json, since otherwise, it may cause problems.

Please mind, that publication on GitHub packages has been discontinued!

Using Cottontail DB as Maven Dependency

You can also use Cottontail DB as Maven dependency, e.g., for use in embedded mode. Just include the following dependency descriptor:

<dependency>
  <groupId>org.vitrivr</groupId>
  <artifactId>cottontaildb</artifactId>
  <version>[version]</version>
</dependency>

To start Cottontail DB in embedded mode, you can invoke CottontailKt.embedded() with a valid Config class instance.