Skip to content

Latest commit

 

History

History
30 lines (22 loc) · 1.8 KB

production-notes.md

File metadata and controls

30 lines (22 loc) · 1.8 KB
Raw

Production Checklist

To ensure MultiJuicer runs as smoothly during your CTF's / trainings / workshops, heres a list of things you might want to make sure is configured correctly before you run MultiJuicer in "production".

  1. Set .balancer.cookie.cookieParserSecret to a random alpha-numeric value (recommended length 24 chars), this value is used to sign cookies. If you don't set this, each helm upgrade you run will generate a new one, which invalidates all user sessions, forcing users to rejoin their team.
  2. As you are running this with https (right?), you should set balancer.cookie.secure to true. This marks the cookie used to associate a browser with a team to transmitted via https only.
  3. Make sure the value you have configured for juiceShop.maxInstances fits your CTF / training / whatever you are running. The default is set to only allow 10 instances. Set to -1 to remove any restrictions.
  4. Set balancer.replicas to at least 2, so that you have at least one fall back JuiceBalancer when one crashes or the node it lives on goes down.
  5. When running a CTF with JuiceShop challenge flags, make sure to change juiceShop.ctfKey from the default. Otherwise users will be able to generate their own flags relatively easily. Additionally, include the juiceShop.nodeEnv value and specify it as "ctf". This way, it will generate flags for the CTF event. The default behavior is to not generate them.

Security Consideration

Add SecurityContext and PodSecurityContext further isolate and secure your training platform.

TLDR

Here's a example helm values file:

balancer:
  replicas: 3
  cookie:
    cookieParserSecret: "THIS_IS_A_EXAMPLE_DONT_USE_THIS_AS_THE_ACTUAL_SECRET"
    secure: true

juiceShop:
  maxInstances: 42
  nodeEnv: "ctf"
  ctfKey: "DONT_LET_ME_FIND_YOU_USING_THIS_EXACT_VALUE"