Modern distributed systems are powerful but their complexity makes them prone to failures that are difficult to recover from.
Temporal.io is an open-source framework that simplifies complex systems and improves reliability through a concept called durable execution.
Applications are built using deterministic workflows for business logic and activities for executing individual, potentially fallible tasks.
Durable execution captures the complete state of a function, allowing it to be seamlessly migrated and resumed on another server after a crash.
The framework includes built-in features like auto-save, native retries, durable timers, schedules, and signals for human-in-the-loop processes.
Developers deploy code to workers that poll the Temporal server for tasks, which in turn captures and stores the complete execution state.
A practical example shows how to define a workflow, create activities, implement worker logic, and use a client to send signals.
The Python SDK provides durable versions of asyncio primitives like timers, tasks, and cancellations that survive process crashes.
The Python SDK is backed by a Rust core and supports multiple activity styles including async, threaded, and multi-process executors for different use cases.
29:32Maxim Fateev
39:04Bobur Umurzokov
33:44Fabian Schindler
46:43Bobur Umurzokov
59:07Shweta Palande
59:43Paul Graham
30:29Iulia Feroli
26:08Maxim Fateev
.webp?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
CONTIAMO GMBH
Berlin, Germany
Peter Park System GmbH
München, Germany
SYSKRON GmbH
Regensburg, Germany
DIGITEC Financial Technologies and Services GmbH
Hamburg, Germany