Overview of our architecture

The architecture at the Neuroenergetics Lab looks like this.

• Davinci is our remote server
  • Once logged in (e.g. with ssh), we can access the Docker engine, Jupyter Hub and our RAID storage
• Bratum is the name of the virtual machine running on Davinci
  • XNAT runs inside Bratum
  • We have the sudo rights within Bratum
• We use the official xnat-docker-compose distribuiton
  • Consists of three containers configured by Docker Compose and running on a separate Docker engine on Bratum:
    • xnat_web is the web application
    • xnat_db is the PostgreSQL database
    • xnat_nginx is the proxy
  • XNAT’s Container Service Plugin has connection to the Docker engine running on Davicni
    • Our (high-performace) Davinci server executes all data processing in Docker containers

XNAT can be accesssed via three different ways. (Click on the corresponding links to connect to xnat now):

• XNAT web. Read about XNAT web here
• ssh connection <username@xnat.tumnic.mgruber.eu>
• Jupyter Hub (requires VPN connection) Set up you VPN Connection here

Note: If you are on Research Network you dont need openVPN for Jupyter Hub connection, otherwise you have to connect.

---