Axum 0.8 WebSocket Heartbeat & Task Desynchronization Invariant

--- description: Axum 0.8 WebSocket Heartbeat & Task Desynchronization Invariant - Enforces deterministic bidirectional heartbeat polling and cooperative Tokio task cancellation for Axum 0.8 WebSocket connections. Eliminates silent file descriptor exhaustion, zombie connection accumulation, and unhandled task panics caused by asymmetric split-stream teardowns during edge reverse proxy failovers and rolling deploys. globs: src/**/*.rs, crates/**/*.rs alwaysApply: true verified_by: "agentrules.org" license: "CC-BY-4.0" source: "https://agentrules.org/rules/axum-08-websocket-heartbeat-desync-guardrail" --- # Axum 0.8 WebSocket Heartbeat & Task Desynchronization Invariant > **Severity**: CRITICAL | **Category**: backend > **Anti-Pattern Warning**: Asymmetric Split Tasks with Detached Ping Loops and Missing Pong Verification ## System Constraints You are a Principal Rust Concurrency Architect enforcing strict lifecycle guarantees for Axum 0.8 WebSocket connections. You must strictly forbid detached, uncoordinated Tokio read/write tasks that lack shared cancellation tokens, monotonic heartbeat deadline tracking, and bounded backpressure. Mandate unified event-loop patterns or coupled cooperative cancellation to prevent zombie connection pools during edge proxy failovers. ## Strict Guidelines - Engineers MUST ALWAYS coordinate WebSocket reader and writer routines via a unified `tokio::select!` loop or linked `tokio_util::sync::CancellationToken` instances. - Engineers MUST ALWAYS verify incoming `Message::Pong` frames and compute monotonic elapsed deadlines (`std::time::Instant`) rather than assuming blind ping delivery guarantees client liveness. - Engineers MUST NOT spawn unbounded, detached Tokio tasks (`tokio::spawn`) from Axum WebSocket handlers without tracking JoinHandles or binding abort guards. - Services MUST intercept application shutdown signals (SIGINT/SIGTERM) and dispatch cooperative `Message::Close` frames with bounded timeouts prior to socket termination. - Components NEVER use unbounded channels (`tokio::sync::mpsc::unbounded_channel`) for outbound WebSocket frame buffering; bounded channels with calibrated capacity MUST be enforced. ## Bad vs Good Implementations ### Naive Detached Read/Write Tasks vs. Unified Select with Monotonic Heartbeat Tracking #### Bad Antipattern (AI Common Hallucination) > **Why it fails**: Splitting the socket into two detached, uncoordinated Tokio tasks leaves the write loop running indefinitely if the client silently drops connection without sending TCP FIN/RST. Pings are emitted blindly with zero verification of client pong reception, accumulating dead tasks, pinned buffers, and leaked file descriptors across rolling deploys. ```rust use axum::extract::ws::{Message, WebSocket}; use futures_util::{SinkExt, StreamExt}; use tokio::time::{interval, Duration}; // ANTI-PATTERN: Detached tasks cause zombie sockets on TCP drop; // unverified pings cause silent FD exhaustion behind reverse proxies. pub async fn handle_socket(socket: WebSocket) { let (mut sender, mut receiver) = socket.split(); // Detached writer: Never terminates if receiver drops or client vanishes tokio::spawn(async move { let mut ping_interval = interval(Duration::from_secs(15)); loop { ping_interval.tick().await; // Blind pinging without checking pong responses let _ = sender.send(Message::Ping(vec![].into())).await; } }); // Detached reader: Blocks forever on dead half-open TCP connections tokio::spawn(async move { while let Some(Ok(msg)) = receiver.next().await { if let Message::Text(text) = msg { println!("Received: {text}"); } } }); } ``` #### Production Standard (Correct) > **Standard**: Unifies stream reading, ping frame emission, and pong verification under a single biased `tokio::select!` block. Enforces a strict monotonic timeout deadline (`CLIENT_TIMEOUT`) that actively severs connections if edge proxies or clients stop acknowledging heartbeats, while gracefully flushing `Message::Close` frames upon SIGTERM reception. ```rust use axum::extract::ws::{Message, WebSocket}; use futures_util::{SinkExt, StreamExt}; use std::time::Instant; use tokio::time::{interval, Duration}; use tokio_util::sync::CancellationToken; const HEARTBEAT_INTERVAL: Duration = Duration::from_secs(15); const CLIENT_TIMEOUT: Duration = Duration::from_secs(35); // PRODUCTION INVARIANT: Coordinated select loop with deadline enforcement & shutdown hooks pub async fn handle_socket(socket: WebSocket, shutdown: CancellationToken) { let (mut sink, mut stream) = socket.split(); let mut ticker = interval(HEARTBEAT_INTERVAL); let mut last_pong = Instant::now(); loop { tokio::select! { biased; _ = shutdown.cancelled() => { let _ = sink.send(Message::Close(None)).await; break; } _ = ticker.tick() => { if last_pong.elapsed() > CLIENT_TIMEOUT { // Client missed heartbeat deadline; abort zombie connection let _ = sink.send(Message::Close(None)).await; break; } if sink.send(Message::Ping(vec![].into())).await.is_err() { break; } } msg = stream.next() => { match msg { Some(Ok(Message::Pong(_))) => { last_pong = Instant::now(); } Some(Ok(Message::Close(_))) | None => break, Some(Ok(Message::Text(payload))) => { let _ = payload; } Some(Err(_)) => break, _ => {} } } } } } ```

Architectural Specification & Invariant Rationale

Architectural Invariant Design: Axum 0.8 WebSocket Heartbeat & Task Desynchronization Invariant

Axum 0.8 WebSocket Heartbeat & Task Desynchronization Invariant defines an industrial-grade behavioral standard for backend engineering. Enforces deterministic bidirectional heartbeat polling and cooperative Tokio task cancellation for Axum 0.8 WebSocket connections. Eliminates silent file descriptor exhaustion, zombie connection accumulation, and unhandled task panics caused by asymmetric split-stream teardowns during edge reverse proxy failovers and rolling deploys. Modern software architectures require strict adherence to deterministic patterns to guarantee high availability, maintainability, and zero security vulnerabilities.

Framework Paradigm Shift & Drift MechanicsLLMs are trained on millions of open-source repositories containing outdated code, legacy idioms, and unpatched antipatterns. When coding assistants generate implementations for this stack, they default to common statistical tokens: Asymmetric Split Tasks with Detached Ping Loops and Missing Pong Verification Without an explicit system instruction file, developers spend hours debugging silent failures.

Failure Modes & Production Blast Radius

Hazard Analysis: What Breaks Without This Rule

The following failure modes have been cataloged from hundreds of hours of production code reviews where AI assistants produced non-compliant implementations:

Failure Mode	AI Symptom (The Anti-Pattern)	Severity	Production Impact & Remedy
AI Hallucination & Deprecation Drift	Asymmetric Split Tasks with Detached Ping Loops and Missing Pong Verification	Critical	Violating this industrial standard introduces regression bugs, security risks, and unexpected runtime failures in production. Enforce explicit behavioral system directives across all AI coding assistants in your repository.
Silent Architecture Inconsistency	AI assistants produce mixed paradigms across files, creating fragmented codebase conventions.	Medium	Increases cognitive load during code reviews and leads to maintenance debt across engineering teams. Pin project conventions in .cursor/rules, CLAUDE.md, and AGENTS.md so all team agents follow identical guardrails.

Architectural Directives in Detail

Step-by-Step Production Directives

Each rule constraint below is injected into the AI agent's system prompt to establish deterministic boundaries:

Engineers MUST ALWAYS coordinate WebSocket reader and writer routines via a unified `tokio

Engineers MUST ALWAYS coordinate WebSocket reader and writer routines via a unified `tokio::select!` loop or linked `tokio_util::sync::CancellationToken` instances.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.