Ok, thanks!

I have a question concerning the WebRTC implementation in your system. I understand that in terms of performance it is significantly better, especially when a user is connected in P2P using STUN. Moreover, this communication is secure, since the information exchanged between the two parties is end-to-end encrypted. But how do you keep track of the actions that the user performs in his WS? From what I understand, it is not possible to audit them nor to define Network Policies that the user must comply with. Can you explain this, please? 🙂

I don't get what you want to achieve

I was just curious from a customer's point of view

what do you mean by "keep track"

Like logs

in v1 we have a solution to log all shell commands in a users workspace

we're currently in the process of replacing webrtc with wireguard, which opens up the door for us to do network policies/auditing and the like

Ohh okay thank you!

it'll be based on tailscale if you've ever heard of that

Oh no I didn't Using this I guess you will be able to have an efficient auditing system without facing again the latency problem

yup yup

Amazing

I'm curious to see this implementation then Because from my understanding, to have an auditing system you need to have a middle man that inspects the traffic, like your previous proxy

But I'm glad you're on the right way to achieve it!

isn't tailscale still WebRTC ?

the auditing is something we'll likely do later, but since the agent in the workspace mitms all of the traffic it's able to report back to coderd

nah it's wireguard

oh yeah no

this is for agent server communication right?

Okay okay

agent server and client agent

alright

i thought it was web related

well, because it used webrtc

that's why I was confused

because you can't just use any protocol from a browser, but since it has nothing to do with browsers then it doesn't even matter

https://github.com/coder/coder/pull/3505 here's the pr @Wepzen

should hopefully be in soon

(blocked on me currently)