fix(backend): break outer loop on client write failure for accesses/TPS

[ayushsingh82]

Summary

client_write_task in backend/src/lib/server.rs sends three kinds of buffered messages to a WebSocket client each iteration: events_buf, accesses_buf, and tps_buf. When send_message() fails for events_buf, the code correctly breaks the outer loop, terminating the task for that (now-dead) connection.

But for accesses_buf and tps_buf, the send happens inside a for loop over the buffered items, and the break on failure only exits that inner for loop — not the outer task loop:

if !accesses_buf.is_empty() {
    for accesses in std::mem::take(&mut accesses_buf) {
        let server_msg = ServerMessage::TopAccesses(accesses);
        if let Err(e) = send_message(&mut ws_sender, server_msg).await {
            error!("Failed to send accesses to {}: {}", addr, e);
            break; // only exits this `for`, outer `loop` keeps going
        }
    }
}

A send_message failure here means the client's WebSocket write is broken (e.g. disconnected). Once that happens, the task should exit — same intent as the events_buf path right above it — but instead it falls through and keeps looping, repeatedly trying (and failing) to serve a dead connection instead of terminating.

Changes

• Label the outer loop as 'outer and change the two nested breaks (in the accesses_buf and tps_buf send loops) to break 'outer, so a write failure on any buffer terminates the task consistently.
• No other behavior changes; the unlabeled breaks elsewhere in the function (broadcast receiver error, events send failure) are unaffected since they weren't nested in a for loop to begin with.

Test plan

• Reproduced the exact control-flow bug in an isolated Rust snippet (bare break inside a for inside a loop only exits the for) and confirmed the labeled-loop fix resolves it.
• cargo fmt --check passes on the crate.
• cargo check / cargo clippy / cargo test — could not run locally on macOS; monad-event-ring's build script requires cmake and Linux hugetlbfs headers not available in this environment (same constraint noted in fix(backend): prevent panic when txn_idx exceeds fixed buffer size #115). Deferred to CI.

Greptile Summary

This PR corrects WebSocket writer termination after TopAccesses or TPS delivery fails.

• Labels the outer client-write loop.
• Exits that loop directly when either nested send loop encounters an error.
• Makes accesses and TPS failure handling consistent with the existing terminal behavior for event-send failures.

Confidence Score: 5/5

The PR appears safe to merge because the changed breaks now terminate work for a connection whose WebSocket writer has already failed.

The labeled breaks correct the nested-loop control flow without changing successful delivery behavior or introducing a reachable lifecycle failure.

Important Files Changed

Filename	Overview
backend/src/lib/server.rs	Correctly changes nested-loop breaks to terminate the client writer after a failed WebSocket write; no actionable regression was identified.

_{Reviews (1): Last reviewed commit: "fix(backend): break outer loop on client..." | Re-trigger Greptile}

[vercel]

@ayushsingh82 is attempting to deploy a commit to the MF Flagship Team on Vercel.

A member of the Team first needs to authorize it.