Durable Execution helps ensure applications continue running despite failures, beyond what defensive coding and hardware redundancy can handle.

The New Stack is a publication covering trends and technologies in cloud-native development, DevOps, and software delivery. Developers can learn about containerization, Kubernetes, and cloud computing, as well as explore topics such as microservices architecture, serverless computing, and continuous integration/continuous delivery (CI/CD) pipelines.

The New Stack

Durable Execution platforms like Temporal ensure applications run to completion despite failures by automatically storing and reconstructing state. Unlike traditional approaches using defensive coding and hardware redundancy, these platforms guarantee execution continues after crashes without duplicating completed steps. The core abstraction is a Workflow that defines task sequences, with non-deterministic operations isolated in Activities. When failures occur, Workers reconstruct state from an Event History log and resume execution. The approach handles transient errors through exponential backoff, supports hotfixing live systems, and eliminates the need for complex infrastructure like message queues and schedulers. Originally developed at Amazon (SWF) and Uber (Cadence), Temporal now powers production systems at Netflix and other enterprises, reducing deployment failures from 4% to 0.0001%.

Durable Execution: Build reliable software in an unreliable world

Reliability through defensive code and hardware redundancy

Rethinking application development with Durable Execution

How Durable Execution handles persistent processing failures

Understanding the Temporal programming model

How Temporal grew from early work at Amazon, Microsoft, and Uber