go-libp2p-transport-upgrader
go-libp2p-transport-upgrader copied to clipboard
Consider dropping connections when the threshold is full
That is, consider dropping connections from the underlying TCP socket. Maybe there's an option we can set on the TCP socket to let the kernel do this?
https://github.com/libp2p/go-libp2p-transport-upgrader/blob/5ddf5deb2271e43bf3c4ffa24841e13c69d4581f/listener.go#L116-L136
This doesn't look like the issue. Instead, it looks like a sleep in the error catcher.
We have a lot of piecemeal mechanisms for backpressure scattered across the codebase. At one point we'll need to consider a reusable backpressure gadget, maybe inspired by SEDA.