[dev] Rust panics after inactivity "Connection reset without closing handshake"

[PierBover]

🐛 Bug Report

I'm creating this bug as requested by @EverlastingBugstopper here: #1047 (comment)

After the computer goes to sleep or after the worker hasn't been used for some time (a couple of minutes) Rust panics and shows this error message:

thread 'main' panicked at 'called `Result::unwrap()` on an `Err` value: Protocol("Connection reset without closing handshake")', src/commands/dev/socket.rs:36:27
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

Environment

• operating system: macOS 10.13.6
• output of node -v: v12.16.1
• output of wrangler -V: ✨ wrangler 1.9.0-rc.0

Steps to reproduce

Open a worker and put the computer to sleep or wait a couple of minutes without using the worker.

[EverlastingBugstopper]

We'll need to figure out if we want to kill the process with a nicer error message here and document that we limit connections to x minutes, if we want to rengoatiate a new websocket connection after a certain period of time, or if we need the preview service to allow longer running sessions.

cc @ispivey - I think the issue here is that the preview service drops the connection after a certain length of time, regardless of whether or not the dev session is sending heartbeat messages.

[PierBover]

rengoatiate a new websocket connection after a certain period of time

This sounds like it would provide the best user experience, no?

What annoys me the most personally is that I need to manually restart the worker very often.

[EverlastingBugstopper]

Yup - I think issuing a reconnect here is probably the best choice :)

[EverlastingBugstopper]

i've opened this issue to see if i can get some help/pointers on how to do this effectively