I know that threat modelling and security sounds like a rather dry topic, but it’s a structured way of thinking about security threats that i feel it’s really important developers start thinking about early in the delivery lifecycle (whichever model you’re using). With containers, it’s more likely that security tools need to be baked or added into container images early in the process so educating developers about this is - in my option - critical for the success of a continuous delivery model. Understanding that security risks to containers are different to virtual machines or the traditional monolithic application stack and that the traditional security toolset isn’t necessarily effective any more is key, and one of the main threads of my talk. As I’m conscious this could be a a dry topic, my talk has many GIFs and memes. More seriously though, it includes relevant anecdotes and was developed with input from Python developers in both my professional and personal circle, to bring (where applicable) python-specific container security examples into the discussion.
A brief summary of the talk structure:
- Introduction to myself and the talk
- Overview of threat modelling
- Threat modelling for containers
- The difference between threats to containers, VMs and the traditional application stack
- Key security threats to containers in general
- Key security threats to containers running Python
- Tools to mitigate these threats
- Where these tools fit within the process
- Summary and close