feat(l1): log warning if no CL messages are received in a while

[SDartayet]

Motivation

Making it clearer when the client isn't working properly

Description

This PR adds an extra thread which keeps track of whether a message from the consensus layer hasn't been received in the past 30 seconds. This is done by having the auth RPC handler communicate with it when a new message is received to reset the countdown. If no messages have been received after 30 seconds, a message is printed to warn the user, along with some hints as to possible reasons why Ethrex isn't detecting the consensus client.

Closes #5234

[github-actions]

Lines of code report

Total lines added: 13
Total lines removed: 0
Total lines changed: 13

Detailed view

+-------------------------------------+-------+------+
| File                                | Lines | Diff |
+-------------------------------------+-------+------+
| ethrex/crates/networking/rpc/rpc.rs | 848   | +13  |
+-------------------------------------+-------+------+

[SDartayet]

I wasn't sure whether to make this warn or error level. I figured error was more appropriate for end user scenarios, since Ethrex would always need to run alongside a consensus client, but it could get cumbersome in testing scenarios where the client is run with no consensus client.

[mpaulucci]

warn is better

[Copilot]

Pull Request Overview

This PR adds a watchdog mechanism to detect and log warnings when no messages from the consensus layer are received within 30 seconds, helping users identify connectivity issues between Ethrex and the consensus client.

• Added a watch channel-based timer that resets on each auth RPC request
• Spawned a monitoring task that logs an error with troubleshooting hints after 30 seconds of inactivity
• Modified the auth RPC handler to send notifications through the watch channel

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The closure signature is incompatible with axum's handler requirements. The async move should come before the pipe characters. The correct syntax is move |ctx, auth, body| async move { ... } to create a closure that returns a future, rather than an async closure.

Suggested change

	let authrpc_handler = async move |ctx, auth, body| {
	let authrpc_handler = move |ctx, auth, body| async move {

[Copilot]

Multi-line string literals should use proper formatting. Consider using a raw string literal or concatenation to avoid the leading whitespace on line 307 which will appear in the logged error message.

Suggested change

	error!("No messages from the consensus layer. Is the consensus client running?
	Check the auth JWT coincides with the one used by the CL and the auth RPC address and port used by it and Ethrex match."
	error!(
	"No messages from the consensus layer. Is the consensus client running?\n\
	Check the auth JWT coincides with the one used by the CL and the auth RPC address and port used by it and Ethrex match."

[Copilot]

The spawned task will panic if the sender is dropped. When timer_receiver.changed() returns an error (which occurs when all senders are dropped), the code logs an error but continues looping. This will cause continuous error logging. The loop should break when changed() returns Err, which indicates the channel is closed.

Suggested change

	if result.is_err() {
	error!("No messages from the consensus layer. Is the consensus client running?
	Check the auth JWT coincides with the one used by the CL and the auth RPC address and port used by it and Ethrex match."
	);
	match result {
	Ok(Ok(_)) => {
	// Successfully received a change, continue looping.
	}
	Ok(Err(_)) | Err(_) => {
	error!("No messages from the consensus layer. Is the consensus client running?
	Check the auth JWT coincides with the one used by the CL and the auth RPC address and port used by it and Ethrex match."
	);
	break;
	}