If you are now thinking of Salt-SSH, this is not completely wrong, but also not right. What we are talking about here, is an other Transport for Salt, a replacement for ZeroMQ.
Why?
First of all, Upstream is interested in this and this is a strong argument. Also we would like to play with the idea of a master that is establishing a connection to a minion and not the other way around like the ZeroMQ transport is doing it. This would be the answer to the use-cases where we can't keep the connection to the minion all the time or establishing a connection to the master is simply not possible because of network configurations - but it would be possible for the master to reach the minion.
How?
- Deep diving into how a SaltStack transport works.
- Getting familiar with SSH and the Python libs available.
- Start implementation.
This project is part of:
Hack Week 17
Activity
Comments
Similar Projects
(Finish) Implementing SSH in Zig by lmulling
Description
Following Zig's philosophy of reinventing the wheel -- for the better, a while ago I've started implementing the ssh protocol in it. I've got as far as implementing all of the primitives, keys, certs, and most of the agent protocol -- what I needed at the time. Now, the aim is to finish the implementation.
Current implementation: git.sr.ht/~mulling/zssh
Goals
- Have a working implementation of the ssh protocol in Zig.
- Be flexible, as to allow for hacking of the protocol (i.e. testing PQC algorithms).
- Be agnostic of cryptography libraries (i.e. libcrypto, leancrypto).
Resources