Execution retries
Automatic function execution retries ensure that failed functions are retried automatically when they encounter an error. This feature is essential for distributed applications, where function calls often depend on other services, APIs, or endpoints that may be slow, flaky, or temporarily unavailable.
In a distributed application, failures are inevitable due to factors like network instability or service downtime. Without a retry mechanism:
- A portion of your workflows may fail, disrupting user experiences.
- Developers must write and maintain complex custom logic to handle retries.
Resonate addresses these challenges by automating retries with minimal setup, improving reliability and reducing the development burden.
You can define a Retry Policy, and Resonate will handle retries for you, ensuring seamless execution even in the face of temporary failures. Resonate automatically applies these retries to every function within the application’s Call Graph, handling errors, exceptions, or failures with ease.
Retry policies are flexible and can be configured at the Application Node level or for individual functions, giving you fine-grained control over how your application handles transient issues.
With a predefined retry policy, Resonate ensures that:
- Transient issues are retried based on your specifications (e.g., backoff intervals, retry limits).
- Your application logic remains clean and focused without additional error-handling boilerplate.
- Failed functions across distributed services or APIs retry automatically until they succeed or meet the retry policy limit.
When you use Resonate to build applications, every function in the application's Call Graph benefits from automatic retries should the execution fail or return an error or exception.
You can configure a Retry Policy at the Application Node level or per function.
Check out examples for the following SDKs:
See Resonate Retry Policies for a deep dive into the built-in Retry Policies and their behavior.
See Call Graph recovery to learn how Resonate can recover from failures at the platform level (after a process crash).