Docker

You can find dstack for Docker on DockerHub here - https://hub.docker.com/r/dstackai/dstack​

Docker Pull Command

docker pull dstackai/dstack

The Docker image can be used to host the enterprise version of dstack.ai on own servers or in own AWS/Azure/Google Cloud account.

You can run dstack.ai in Docker using the following command:

docker run -d --name <dstack container name> \
-v <path to data directory>:/data \
-p <port on host>:8080 -e dstackai_port=<public port> \
-e dstack_smtp_host=<SMTP server host> \
-e dstack_smtp_port=<SMTP server port> \
-e dstack_smtp_user=<SMTP server user> \
-e dstack_smtp_password=<SMTP server password> \
-e dstack_smtp_starttls=true \
-e dstack_smtp_from=<admin email address> \
dstackai/dstack:latest

By default, it uses the built-in SQLite database to store data. If you want, you can connect the Docker image to your existing Cassandra cluster. To do that, use the following environment variables:

-e dstack_cassandra_hosts=<comma-separated list of hosts> -e dstack_cassandra_port=<port> -e dstack_cassandra_user=<user> -e dstack_cassandra_password=<password>

In this case the Docker image will use the dstack keyspace. If it doesn't exist, it will create it. The Docker image handles schema evolution automatically.

If you'd like to test how the Docker image works without configuring an SMTP server, you can try to run it via the docker-compose.yml.

In case you don't want to use Docker or the multi-user version of dstack, you also can use github.com/dstackai/dstack to start a dstack server locally via command line.