mirror of
https://github.com/outbackdingo/firezone.git
synced 2026-01-27 18:18:55 +00:00
370a0e4b4e
When a relay restarts, our local credentials will be invalid and no amount of retrying can fix this. Currently, the `Allocation` can end up in busy-looping state if: 1. The allocation would be due for a refresh 2. The relay was restarted 3. We haven't received new credentials yet because we haven't made a new connection attempt that uses the same relay The above was observed in #4521 and results in log-spam of: > Invalid credentials, refusing to re-authenticate refresh One part of the state machine correctly discards the message instead of re-sending it. Unfortunately, the result of (1) means there is still a timer that fires and attempts to refresh the allocation. To stop this busy-looping behaviour, we need to invalidate the allocation if we detect that our credentials are wrong. This will also invalidate the candidates which will fail any connection that is currently using this relay. This would have likely already happened before because a relay that is restarted would have lost all channel bindings and thus, the ICE timeout will kick-in.
Rust development guide
Firezone uses Rust for all data plane components. This directory contains the Linux and Windows clients, and low-level networking implementations related to STUN/TURN.
We target the last stable release of Rust using rust-toolchain.toml.
If you are using rustup, that is automatically handled for you.
Otherwise, ensure you have the latest stable version of Rust installed.