Skip to main content

NotificationRetryBehavior

Namespace: RA.Utilities.Feature.Behaviors

NotificationRetryBehavior is an implementation of the INotificationBehavior interface. Its primary purpose is to add resilience to your event handling system by automatically retrying failed notification handlers.

When a notification is published, it's processed by its handlers. Sometimes, a handler might fail due to a transient issue, such as a temporary network problem, a database deadlock, or a brief service unavailability. Instead of immediately failing and logging an error, this behavior intercepts the exception and retries the operation.

🛠️ How it works:​

  1. Intercepts Execution: It wraps the call to your INotificationHandler(s) (represented by the next delegate) inside a try-catch block within a loop.

  2. Retries on Failure: If an exception occurs, the catch block is executed. It logs that an attempt failed and then waits for a short, increasing duration (a linear backoff strategy: 200ms, 400ms, etc.).

  3. Sets a Limit: It will retry up to a maximum number of times (hard-coded to 3). If all retries fail, the final exception is allowed to propagate up the stack.

  4. Succeeds on Pass: If any attempt succeeds, the loop breaks, and the process continues normally.

This behavior is crucial for building robust, self-healing systems, as it can automatically recover from temporary glitches without manual intervention.