Circuit Breaker Core Concepts

A circuit breaker is a design pattern that helps you build resilient distributed systems by preventing repeated failures from overwhelming your application. It works by monitoring interactions with remote services or components and temporarily blocking requests when problems are detected. The circuit breaker manages this process through three main states: closed, open, and half-open.

Circuit Breaker States

• Closed state: all requests are allowed to pass through to the remote service;
• Open state: requests are blocked and fail immediately, without contacting the remote service;
• Half-open state: a limited number of requests are allowed to test if the remote service has recovered.

Example: Service-to-Service Communication

Imagine a payment service that calls an external credit card processing API. When the API is healthy, the circuit breaker is in the closed state and all payment requests go through. If the API starts failing (for example, due to network issues), the circuit breaker moves to the open state and blocks new payment requests, preventing the payment service from being overloaded by repeated failures. After a short period, the circuit breaker enters the half-open state and allows a few test requests to see if the API is working again. If those succeed, the circuit breaker returns to the closed state; if they fail, it goes back to open.

Failure Thresholds

A failure threshold determines how many errors must occur before the circuit breaker trips from closed to open.

• Thresholds can be based on: a count of consecutive failures; a percentage of failed requests over a time window; or a combination of both.

Example

If you set a failure threshold of 5, the circuit breaker will open after 5 consecutive failed payment requests to the API. This prevents your application from repeatedly attempting operations that are likely to fail.

Timeout Policies

A timeout policy defines how long the circuit breaker waits before moving from the open state to the half-open state to test if the service has recovered.

• Timeouts help avoid: overwhelming a failing service with too many requests;
• Timeout duration: should be long enough to give the remote service time to recover, but short enough to avoid unnecessary delays.

Example

You configure the circuit breaker to stay open for 60 seconds after tripping. After this timeout, it transitions to half-open and allows a few test requests.

Recovery Strategies

Recovery strategies determine how the circuit breaker returns to normal operation:

• Success-based recovery: if test requests in the half-open state succeed, the circuit breaker closes and resumes normal traffic;
• Failure-based fallback: if test requests fail, the circuit breaker reopens and waits for another timeout period.

Example

After the timeout, the payment service sends 2 test requests. If both succeed, the circuit breaker closes and resumes processing all payments. If either fails, it reopens and waits before trying again.

Circuit breakers help you build stable, reliable systems by automatically handling failures, preventing cascading problems, and allowing services to recover gracefully.