This article reviews seven widely used Python web frameworks—Django, Flask, Scrapy, Tornado, Web2py, Weppy, and Bottle—detailing their main features, advantages, and drawbacks to help developers choose the most suitable tool for their projects.
This article examines the challenges of deploying dozens of deep‑learning models at Zuoyebang and compares three serving architectures—Gunicorn + Flask + Transformers, Tornado + PyTorch, and Tornado + Triton—highlighting performance trade‑offs and presenting a final high‑concurrency, low‑latency solution in production.
This article introduces seven open‑source Python web frameworks—Django, Flask, Scrapy, Tornado, Web2py, Weppy, and Bottle—detailing their main features, typical use cases, and the key advantages and disadvantages of each to help developers choose the most suitable framework for their projects.
The article presents DuLab, a distributed UI‑automation platform for mobile apps that leverages Airtest, Python/Tornado, Celery, and Redis to schedule and run bulk test cases on remote devices, detailing its layered case management, package database, mock services, and the resulting scalable, efficient regression testing across diverse devices.
This article continues a series on server‑side template injection by presenting four hands‑on labs covering Tornado (Python), Velocity (Java), Freemarker (Java) and a Freemarker sandbox‑escape, detailing syntax basics, attack surfaces, exploit payloads, defensive measures, and step‑by‑step exercises.
This article demonstrates how to build a basic Tornado server, measures its performance under blocking operations, and then progressively applies multithreading, thread‑pool executors, and finally asyncio to achieve dramatically higher QPS and lower response times.
This article explains the fundamentals of asynchronous programming, contrasting blocking, non‑blocking, synchronous and asynchronous execution, and demonstrates how to implement async behavior in Python using Tornado, asyncio, and aiohttp, including coroutine definitions, yield usage, and practical code examples.
This article provides a detailed examination of Tornado’s IOLoop core, explaining its callback handling, timer management, I/O event loop, and the pivotal role of Future objects, while also demystifying the implementation of gen.coroutine and offering practical code examples to illustrate asynchronous execution.
The article details the design, architecture, installation, self‑updating mechanism, concurrent task handling, and multi‑layer monitoring of a custom Python‑based HttpAgent used in a complex test environment, highlighting its advantages over traditional SSH execution.
This article compares Python web deployment options, explaining the concepts and workflows of CGI, FastCGI, WSGI, uWSGI, Gunicorn, Tornado, and related tools, and offers guidance on selecting the appropriate server and protocol for production environments.
This article explains the inner workings of Tornado's IOLoop, covering its three core components—callback handling, timeout scheduling, and I/O event processing—while also detailing the Future class and the gen.coroutine implementation that together enable asynchronous programming in Python.
This tutorial explains how Python decorators work, demonstrates a simple decorator that adds database‑connection logic, shows how to return a wrapper function to avoid modifying existing calls, and walks through creating a Flask‑style @router decorator for Tornado web routing.
This article recounts Zhihu's evolution from a two‑person startup to a platform serving over 80 million monthly users, detailing early Python/Tornado choices, high‑availability, custom logging, event‑driven design, rendering optimizations, and the transition to a service‑oriented architecture.
This article reviews various web real‑time message push techniques—including polling, long‑polling, iframe streaming, socket plugins, and WebSocket—evaluates their pros and cons, and explains why the author selected the Comet approach implemented with the Python Tornado framework for a web push system.
The article chronicles Zhihu's growth from a two‑engineer startup using Python and Tornado on a single Linode server to a large‑scale backend employing high‑availability MySQL, Redis sharding, a custom distributed logging system (Kids), event‑driven processing with Sink and Beanstalkd, component‑based page rendering via ZhihuNode, and a multi‑layer SOA built on evolving RPC frameworks.
This article describes how a Linux operations team streamlined frequent web code releases by building a web‑based deployment system that combines Git for version control, Rsync for synchronization, and Python/Tornado for backend services, eliminating manual intervention and improving efficiency.
