Description
The decision of which remote computers you can trust and which you can’t is hard. You probably think that you can trust a machine that you’ve just done a fresh install upon, but when that machine leaves your physical control and is shipped off somewhere in the world, has it then left your trust boundary and is completely untrusted? How can you know it’s the same machine, running the same configuration, and it hasn’t been tampered with?
In this talk I’ll discuss techniques that can be used to try to keep that machine in a trustworthy state, and procedures to prove that it’s still running the expected configuration and code.
The same and closely related techniques are also used to prevent the machine from running code which aren't approved.
Another side of the same techniques can be used to establish trust in a remote machine that’s not in your physical control, like a VM in a cloud provider, a remote machine needing a reinstall, initial keying of your fleet of compute nodes etc.
This talk is about TTP’s which are applicable to any system where you run reasonably controlled workloads, like managed laptops, workstations, kubelets, or as in my case, Linux appliances.
| Length | 45 minutes |
|---|