Circuit Breaker — Closed, Open & Half-Open

A failing provider, a live request stream, and the breaker's state machine. Without it, every call waits the full timeout then fails. With it, repeated failures trip the breaker OPEN (calls fast-fail to fallback), a cooldown ticks, then HALF-OPEN probes decide whether to close again. Watch the states flip in real time.

The flight provider is failing. Watch the breaker trip OPEN, fast-fail every call to a fallback, then probe in HALF-OPEN. Flip the provider back to healthy and watch a probe CLOSE the circuit automatically.

CLOSED

calls flow to the provider

0/4 fails

→

OPEN

calls fast-fail to fallback

→

HALF-OPEN

one probe call is allowed

Closed —4 fails→ Open —5 ticks→ Half-Open —probe ok→ Closed (probe fails→ Open)

Request stream (newest →)

🟢 success🔴 provider failed⚪ short-circuited (fallback)🟡 probe

Breaker state

CLOSED

Provider calls

Fast-failed (saved)

calls the breaker spared the provider

State transitions

// watching for failures…

Watch with the breaker OFF first: a failing provider gets called on every request (look at "Provider calls" climb). Turn it ON and the breaker trips after 4 failures — calls stop hitting the provider and fast-fail instead.

What just happened

▹A circuit breaker watches calls to a dependency. While it's CLOSED, calls flow normally. After a run of failures it trips OPEN and stops calling — every request fast-fails to a fallback instead of waiting on a service you already know is broken.
▹After a cooldown it goes HALF-OPEN and lets a single probe through. If the probe succeeds it CLOSES (back to normal); if it fails it re-OPENs for another cooldown. That's the whole state machine: Closed → Open → Half-Open → Closed.
▹The point is resource protection: without a breaker every request waits the full timeout on a dead provider, piling up threads and latency. The breaker turns slow cascading failure into instant, cheap fallbacks — and probes for recovery on its own.