Resilience Patterns

JustAPI incorporates enterprise resilience primitives directly in the Rust engine.

Automatically stop calling a failing service when errors exceed a threshold:

app.enable_circuit_breaker(
    failure_threshold=5,      # Open after 5 consecutive failures
    reset_timeout_ms=30000,   # Try again after 30 seconds
)

States:

• Closed — Normal operation, requests pass through
• Open — Requests fail fast with 503
• Half-Open — Probe request allowed; success → Closed, failure → Open

Deduplicate concurrent identical requests so only one hits the backend:

app.enable_request_coalescing(
    headers=["accept"],  # Consider these headers for dedup keys
)

When 100 identical requests arrive simultaneously:

• Only 1 request reaches the handler
• The response is shared across all 100 callers
• Saves backend resources on cacheable/read-only endpoints

app.enable_rate_limiter(max_requests=1000, window_seconds=60)

app.enable_ip_rate_limiter(max_requests=100, window_seconds=60)

Exceeding limits returns 429 Too Many Requests with Retry-After header.

Limit concurrent requests to prevent resource exhaustion:

app.enable_bulkhead(max_concurrent=100)

When the limit is exceeded, requests receive 503 Service Unavailable.

Cancel requests that exceed the time limit:

app.enable_timeout(seconds=30)

Returns 504 Gateway Timeout when exceeded.

app.enable_fallback_middleware(
    fallback_response={"status": "degraded", "message": "Please try again later"},
)

JustAPI handles SIGINT/SIGTERM automatically:

1. Stop accepting new connections
2. Drain in-flight requests (with configurable timeout)
3. Call plugin on_shutdown() hooks
4. Close database connections
5. Exit cleanly

In production, Rust panics abort the process (not unwind). The supervisor (Docker, systemd, K8s) restarts the process immediately:

[profile.release]
panic = "abort"

This avoids undefined behavior across the PyO3 FFI boundary during panics.

app = JustAPIApp()

# Resilience
app.enable_circuit_breaker(failure_threshold=5, reset_timeout_ms=30000)
app.enable_request_coalescing(headers=["accept"])
app.enable_rate_limiter(max_requests=1000, window_seconds=60)
app.enable_ip_rate_limiter(max_requests=100, window_seconds=60)
app.enable_bulkhead(max_concurrent=100)
app.enable_timeout(seconds=30)

• Production Checklist — Production deployment
• Security — Security policy and OWASP compliance
• Observability — Monitoring resilience signals