Resilience Patterns

The Retry with Backoff pattern is a resilience strategy that automatically retries a failed operation after waiting for an increasing amount of time. This is particularly useful when dealing with transient failures, such as network glitches, temporary service unavailability, or resource contention. The …

Read More

Load shedding is a mechanism for gracefully handling overload situations in a system. When demand exceeds available resources (CPU, memory, network, database connections, etc.), a load shedding strategy is employed to selectively reject or delay requests, preventing the system from complete failure. It prioritizes …

Read More

The Idempotent Receiver is a pattern used in distributed systems to ensure that a receiver can safely reprocess the same message multiple times without causing unintended side effects. This is crucial in scenarios where message delivery is not guaranteed – a message might be retried due to network issues or receiver …

Read More

Rate limiting is a technique used to control the rate at which users or systems can access a resource. This is crucial for preventing abuse, ensuring service availability, and protecting backend systems from overload. By limiting the number of requests within a specific timeframe, rate limiting helps to maintain a …

Read More

Sacrificial Architecture is a resilience pattern where non-critical components are intentionally designed as single points of failure, allowing them to be quickly replaced or rebuilt in the event of an outage. The goal isn’t to make these components reliable, but to make their failure cheap and fast, protecting …

Read More

The Bulkhead pattern isolates parts of an application from each other, preventing failures in one part from cascading and bringing down the entire system. It’s inspired by the compartmentalization of a ship’s hull – if one section is breached, the bulkheads prevent the flooding from spreading to the rest …

Read More

The Bulkhead Cache pattern enhances system resilience by isolating a shared, potentially unreliable resource – a cache – from the rest of the application. It achieves this by limiting the number of concurrent requests that can access the cache. If the cache is slow or fails, the bulkhead prevents cascading failures by …

Read More

The Fallback pattern provides a secondary mechanism to fulfill a request when the primary attempt fails. It’s a key component in building resilient systems, ensuring continued operation even in the face of partial failures. This pattern doesn’t actively prevent failure, but gracefully handles it by …

Read More

The Circuit Breaker pattern protects a system from repeated failures of a dependent service. It monitors calls to the dependency and, upon observing a certain failure rate, “opens the circuit” to prevent further requests from being sent. This allows the dependent service to recover without being …

Read More