Security and authentication workarounds for Qdrant vector database
Qdrant is one of the new generation of databases that allows content to be indexed by vectors — the vectors most often being created by large language model (LLM) AIs. These databases become powerful content stores for LLMs to use as a backend for content.
There are a number of vector database options. Qdrant is just one good option.
Qdrant is available on a commercial basis from Qdrant Cloud where they will host and secure your data.
Qdrant’s core is also available open-source via a Docker image. This is great but this Docker image is provided without any authentication abilities. In contrast the Qdrant Cloud version has API key authentication built in.
So how can you set up Qdrant open-source securely? There are many ways to set it up, here are two simple ways to get started with authentication …
Firewalls
The Qdrant database in the Docker image talks to the outside world using a Rest API on port 6333. Check out the docker image for more information.
If you have an application which calls the Qdrant database for content information and that is your only need, then you can put both within the same docker-compose file and they will be able to talk to each other without you…