This article demonstrates how to embed a hidden Facebook Like button that automatically registers likes as the user moves the mouse, using HTML, CSS positioning, JavaScript event handling, and the Facebook SDK, while keeping the element invisible to the user.
Greg Foster investigates why Facebook abandoned Git for Mercurial, revealing that massive repository size caused severe performance bottlenecks, Git’s inability to scale, and the collaborative culture of Mercurial’s maintainers, ultimately shaping Facebook’s tooling decisions and influencing broader industry practices.
The article explains how Facebook abandoned Git for a heavily customized Mercurial and how Google, facing similar monorepo scaling problems, built its own version‑control system called Piper, illustrating the challenges of managing ultra‑large codebases at internet giants.
This article explains Facebook’s internal configuration management system, detailing how tools like Configerator treat configuration as code, support version control, automated validation, large‑scale distribution, UI editing, and canary testing to prevent errors and streamline deployment across services.
The article explains Facebook’s autonomous integration testing infrastructure, detailing how it abstracts services, employs fuzzing, mock services, and record‑and‑playback techniques to enable scalable, isolated testing of backend microservices while minimizing friction and ensuring reliable software delivery.
In November 2022, Ireland's Data Protection Commission fined Meta €265 million for failing to safeguard personal data of over 5 billion Facebook users, highlighting the EU's intensified enforcement of privacy regulations such as GDPR against major tech firms.
This article analyses seven years of Facebook mobile development data to show how increasing release frequency can coexist with stable engineer productivity and unchanged software quality, using metrics such as daily code changes, commit frequency, bug severity, crash rates, cherry‑picks and launch‑blockers.
This article examines Facebook's mobile continuous deployment approach, detailing how increased release frequency, automated pipelines, and a clear separation between development and release engineering have allowed weekly Android and iOS updates without sacrificing software quality or developer productivity.
The article details Facebook’s multi‑year effort to migrate its heavily customized MySQL 5.6 deployment—including the MyRocks storage engine—to MySQL 8.0, describing the technical challenges, patch‑porting strategy, replication changes, automated verification, and application validation performed during the upgrade.
Facebook experienced a six‑hour worldwide outage that knocked out its main site and services like Instagram, WhatsApp, Messenger and Oculus, and engineers later traced the incident to a misconfigured backbone router that broke BGP routing and DNS resolution, sparking conspiracy rumors about data leaks.
Facebook’s engineering team detailed the multi‑year effort to upgrade a heavily customized MySQL 5.6 deployment to MySQL 8.0, describing the migration strategy, code‑patch classification, row‑based replication adoption, extensive automation testing, and the performance and compatibility issues they resolved along the way.
ByteDance’s multi‑billion‑dollar acquisition of VR maker Pico mirrors Facebook’s Oculus purchase, intensifying a tech rivalry that spans metaverse ambitions, short‑video dominance, and aggressive investments in VR hardware, gaming, and social platforms.
The article explains how Facebook’s Shard Manager tackles large‑scale system sharding by combining stateful and stateless service deployment, consistent hashing versus sharding, fault‑as‑normal principles, replication, automated failover, load‑balancing, and elastic scaling to achieve high availability and efficient resource use.
This article explains how Facebook engineered a massive‑scale continuous delivery pipeline—evolving from manual cherry‑pick releases to a near‑continuous deployment system that safely pushes hundreds of code changes per hour to both web and mobile users worldwide.
Facebook detailed its multi‑year journey from MySQL 5.6 to 8.0, explaining why it stayed on 5.6 for MyRocks development, the massive patch‑porting effort, replica‑set migration steps, adoption of row‑based replication, extensive automation testing, and the performance and compatibility issues uncovered along the way.
Facebook’s engineering team detailed a multi‑year, complex migration from MySQL 5.6 to 8.0, highlighting custom‑feature porting, replication compatibility, query changes, performance regressions, and documentation gaps, and explaining how they finally achieved a successful upgrade.
The Supreme Court ruled APIs uncopyrightable, letting Google use Java code in Android, while LG exits the smartphone market, a massive Facebook data leak surfaces, UK developer jobs surge amid skill shortages, and Python 3.9.4 patches critical security flaws.
This article explains how Facebook’s Twine system orchestrates containers across millions of servers, detailing its architecture, support for stateful services, cross‑data‑center control, elastic capacity handling, and the lessons learned from eight years of large‑scale operations.
Facebook’s 2020 open‑source report reveals a rapid expansion to over 700 repositories and 1.3 million GitHub followers, highlights AI breakthroughs like the M2M‑100 multilingual model and ReBel algorithm, and showcases frontend tools such as React Native and the newly released Recoil library.
This article explains how Facebook built a highly scalable continuous delivery pipeline for both web and mobile products, evolving from simple trunk‑based releases with cherry‑picks to a quasi‑continuous deployment model that supports thousands of engineers, rapid rollouts, and robust quality monitoring.
Facebook’s open‑source Pysa tool statically scans Python code to detect data‑flow vulnerabilities, XSS and SQL‑injection risks, leveraging Pyre and Zoncolan techniques, achieving rapid analysis of millions of lines and uncovering 44% of Instagram’s security flaws in early 2020.
This paper describes Facebook's continuous deployment practice for its Android and iOS mobile apps, outlines the release workflow, presents a seven‑year data set covering commits, crashes, and issue tracking, and analyses how deployment frequency, team size, and release timing affect developer productivity and software quality.
The article explains how Facebook uses the GateKeeper configuration system to perform safe, incremental rolling releases of new features, describing the release stages, constraint‑based switch logic, integration with external data stores, and the trade‑offs between flexibility and operational risk.
The article outlines Presto’s high‑performance, coordinator‑worker architecture and query flow, describes YouZan’s migration from mixed Hadoop deployment to dedicated low‑latency clusters, details challenges such as small‑file handling and regex backtracking with their fixes, and previews future enhancements like Alluxio integration, session property managers, and Ranger‑based multi‑tenant isolation.
In 2019, IDCF organized three in‑depth DevOps case study events where participants analyzed the practices of leading companies such as Facebook, Microsoft, Etsy and others, highlighting hacker culture, automated Windows builds, and a five‑step tech transformation roadmap.
This case study examines Facebook's origins, mission, core leadership, explosive growth, business model, and especially its distinctive engineering culture—known as The Hacker Way—highlighting practices such as bootcamps, hackathons, hackamonths, information transparency, and empowerment that drive innovation and align employee performance with company goals.
In this interview, former Facebook Tech Lead Ge Jun shares how rapid yet sustainable development, trust‑based culture, and practical DevOps practices boost R&D efficiency, offering concrete process, engineering, personal, and management strategies along with measurable metrics for tech teams.
Mark Zuckerberg reflects on Facebook’s tumultuous 2018, detailing crises from data scandals to advertising missteps, while analyzing how the company’s vision, ad innovations, and five key success factors have shaped its business model and future challenges.
At F8 2023, Facebook unveiled a “Private Vision” emphasizing encrypted data and limited content visibility, introduced a “secret crush” dating tool, rolled out the ultra‑lightweight Lightspeed Messenger, highlighted community‑centric Groups and cross‑app integration, and signaled major AR/VR/AI investments while showcasing research on user well‑being and new business‑enablement capabilities for developers.
The article examines whether Facebook can reinvent its business model to emulate WeChat’s all‑in‑one digital ecosystem, outlining required shifts in messaging integration, in‑app commerce, and payment systems while highlighting cultural, regulatory, and competitive obstacles that make such a transformation exceedingly difficult.
Facebook faces a wave of legal and regulatory scrutiny after the DC Attorney General sued the company over alleged privacy violations tied to the Cambridge Analytica scandal, while new reports reveal extensive data‑sharing agreements with over 150 tech firms and multiple recent data breaches affecting tens of millions of users.
This article explains how React Native has become the leading cross‑platform mobile framework, details its current use at Facebook, outlines architectural challenges, and describes the major refactoring effort announced by Sophie Alpert to make the framework lighter, more flexible, and better integrated with native code.
Jason Fried’s multi‑year effort to migrate Facebook’s massive codebase from Python 2.7 to Python 3 illustrates the technical, cultural, and organizational challenges of large‑scale language adoption and the concrete strategies that finally made Python 3 the default at the company.
Facebook upgraded Messenger’s storage by redesigning the schema, switching from HBase to the MyRocks MySQL engine, and moving to flash storage, achieving seamless migration without downtime, cutting storage use by 90%, reducing latency fiftyfold, and simplifying operations for billions of users.
Facebook suspended data‑analysis firm CubeYou over a scandal involving misuse of millions of users’ personal quiz data, while NASA launched a crowdsourced project and MIT‑trained AI to scan WISE images, achieving 97% agreement with humans and uncovering 367 new exoplanet‑candidate debris disks.
The Facebook‑Cambridge Analytica scandal exposed the illicit harvesting of 50 million users' data, political manipulation allegations, multiple investigations, a sharp stock plunge, executive resignations, and heightened global scrutiny of data‑privacy practices.
This article presents a curated list of Facebook’s most popular open‑source projects, ordered by GitHub star count, covering a wide range of technologies from front‑end libraries like React to big‑data engines such as Presto, providing concise descriptions and links for each.
Since its inception, Facebook has relied on MySQL to handle data from over two billion users, but recent shifts toward NoSQL have led to the development of Apollo—a layered storage system inspired by Paxos, Raft, RocksDB, and custom APIs, aiming to improve scalability, latency, and fault tolerance.
Facebook announced it will relicense React, Jest, Flow, and Immutable.js to MIT, sparking backlash from the developer community and prompting companies like Apache, WordPress, and Baidu to reconsider their use of React, while the company explains its reasoning and future plans.
Facebook’s architecture, built on a LAMP‑style foundation yet vastly scaled, relies on technologies such as Memcached, HipHop for PHP, Haystack, BigPipe, Cassandra, Scribe, Hadoop, Hive, Thrift, and Varnish, alongside progressive deployment strategies like Gatekeeper, to handle billions of daily requests and petabytes of data.
This article explores whether Facebook could be hosted on AWS by estimating Facebook's server count, comparing hardware performance, evaluating software compatibility, and calculating the massive cost implications, ultimately concluding that while technically possible, it would be far more expensive than maintaining its own data centers.
This article explores the cutting‑edge data pipelines of Netflix, Facebook, Airbnb and Pinterest, detailing the massive event volumes they handle, the core technologies such as Kafka, Spark, Presto and Hadoop, and how these giants design scalable, real‑time analytics infrastructures.
This article traces Facebook's evolution from a small social site to a global platform, explains how its massive data‑storage challenges led to the adoption of NoSQL solutions like Cassandra and HBase, and breaks down the core patterns, consistency models, and scaling techniques that power such large‑scale systems.
This article explains why interest‑driven feeds are essential, reviews Facebook's NewsFeed evolution and EdgeRank algorithm, details Pinterest's Smart Feed architecture and Pinnability model, and provides a comprehensive guide to building, ranking, and monitoring generic interest‑feed systems for social platforms.
This article shares a former Facebook engineer’s deep dive into the company’s 12‑year evolution, revealing how hack culture, a balanced design‑engineering approach, open data policies, OKR‑driven performance reviews, and strategic talent management have shaped a world‑class engineering organization and offer lessons for Chinese tech firms.
This article explains Facebook’s BigPipe architecture, which breaks pages into independent pagelets and pipelines server‑side generation with client‑side rendering to dramatically cut perceived latency and double overall site speed.
The article analyses Facebook’s F8 conference, explaining how its focus on live‑video APIs, 360° video, VR/AR, AI‑powered Messenger bots and Free Basics reflects a strategic push to deepen the Facebook platform as the core organizer of future digital interactions.
During an exclusive visit to Facebook’s Menlo Park campus, the author uncovers the company’s sophisticated release engineering practices—including the HipHop optimizer, a custom BitTorrent‑based deployment system, continuous testing, and a unique “Hotfix Bar” culture—revealing how billions of daily requests are reliably delivered at massive scale.
Facebook’s Dragon distributed graph query engine optimizes high‑volume single‑hop and multi‑hop queries by introducing specialized indexing, socially aware inverted indices, and functional primitives, dramatically reducing latency, storage reads, and CPU usage while handling massive social‑graph data.
This article shares insights from former Facebook engineers on the company’s engineering culture, open workspace, code‑review practices, and the technical architecture behind the News Feed, including real‑time publishing, push/pull models, and machine‑learning‑driven ranking.
Facebook tackled Android performance monitoring by replacing manual instrumentation with a rule‑based bytecode rewriter that automatically injects tracing markers, minimizes overhead, and captures cross‑thread activity, offering a scalable solution for real‑world telemetry collection.
Facebook’s prolific open‑source strategy, driven by a culture that encourages engineers to build experimental tools like React and GraphQL, illustrates how the company balances risk, internal opposition, and talent attraction to turn many side projects into influential technologies.
Using publicly released H1‑B visa data from 2012‑2014, this analysis compares base salaries of software engineers and product managers at Google and Facebook, revealing similar pay ranges, notable salary peaks, and highlighting data limitations and potential biases affecting the conclusions.
This article explains Facebook's large‑scale MySQL backup architecture, covering the Python‑based automation framework, master‑slave deployment, logical mysqldump backups, warm and cold storage locations, source selection heuristics, full and incremental backup pipelines, verification processes, and future RBR‑based improvements.
An experienced Facebook engineer explains that promotions rely on sustained technical excellence, measurable impact, peer feedback, and a transparent calibration process rather than office politics, outlining the review steps, criteria, and the importance of helping others.
This article details Facebook's MySQL backup architecture, covering preparation, logical backup format, storage locations, source selection, full and incremental backup pipelines, verification mechanisms, and future directions such as RBR‑based logical incremental backups.
The article examines Mark Zuckerberg's multi‑year vision for Facebook, detailing its strategic focus on artificial intelligence, virtual reality, and ambitious connectivity projects such as Internet.org and drone‑based internet, while recounting key acquisitions, internal culture, and the company's massive user growth.
The article explores Facebook’s experimental AI‑driven virtual assistant M, detailing its four‑step workflow, the role of human coaches, real‑world use cases like flower ordering and ticket booking, and the broader business implications for commerce and competition with Google.
This article shares Facebook’s four design principles for B2B products—helping users grow, balancing efficiency with effectiveness, turning complexity into simplicity, and delivering precise predictions—illustrated with real‑world examples, A/B test results, and the evolution of their advertising platform.
The article chronicles Facebook's evolution from a social platform into a leading AI research hub, detailing the founding of FAIR, its key personnel, ambitious goals, major projects such as memory networks, embedding world, DeepFace, language technology, and the M assistant, and highlights the open, collaborative nature of its AI work.
In a recent Menlo Park Q&A, Mark Zuckerberg outlined four forward‑looking ideas—including a new emotion reaction button, AI‑based image filtering, a video‑centric future with virtual reality, and the candid admission that Facebook, like any company, won’t last forever.
Facebook’s web team revisited Yahoo’s 2007 cache research, measuring how often PC and mobile browsers serve resources from cache today, discovering that roughly 45% of requests still miss the cache and that cache lifetimes are short, prompting new optimization recommendations.
This article explains how Facebook tackles massive recommendation data—over 100 billion ratings—by using distributed collaborative filtering, matrix factorization, SGD/ALS hybrid algorithms, and a novel work‑to‑work communication scheme built on Apache Giraph to achieve high performance and scalability.
The article explains how Facebook built a large‑scale distributed recommendation system using Apache Giraph, collaborative filtering with matrix factorization, SGD and ALS algorithms, a novel work‑to‑work communication scheme, and performance optimizations that achieve ten‑fold speedups on billions of ratings.
Facebook’s data warehouse, storing over 300 PB and growing by 600 TB daily, transitioned from the RCFile format to an optimized ORCFile implementation, achieving 5‑8× better compression and up to three‑fold faster write performance while maintaining high read efficiency.
The article examines how Facebook’s massive data centers operate at scale, detailing server serviceability, automated fault‑diagnosis systems like CYBORG, staffing ratios, and the engineering practices that enable reliable service for billions of users worldwide.
Facebook’s Cold Storage system redesigns hardware and software to store rarely accessed data with up to 75% lower power consumption, modular 2U racks holding 30 disks, Reed‑Solomon erasure coding for cheap redundancy, and a self‑healing “anti‑entropy” process that improves performance as the system scales.
This article examines Facebook's cold storage system, detailing how the company redesigned hardware and software to slash power consumption, improve reliability with Reed‑Solomon coding, mitigate bit‑rot, and balance loads while supporting massive photo archives in energy‑constrained data centers.
The talk explains Facebook's early engineering culture—emphasizing rapid "Hack" iterations, a decentralized organization that avoids dependencies, open code‑base ownership, data‑driven debate, and the strategic hiring of strong engineers—to illustrate how these practices drive high‑speed product development.
The article analyzes Facebook's massive image storage challenges and explains the Haystack architecture, detailing its components—Directory, Store, and Cache—how it reduces I/O, manages metadata, and handles read/write operations at billions‑scale while also addressing CDN dependency and fault tolerance.
The article breaks down Facebook's multi‑layered architecture, explaining how its PHP‑based web tier, Thrift‑driven services, Memcached caching, and specialized Erlang and C++ servers work together to deliver the News Feed and real‑time chat at massive scale.
The article details Facebook’s LAMP‑based architecture, describing how HipHop compiles PHP to C++, Thrift‑based services in PHP, C++, and Java run on custom servers, and how MySQL, Memcached, Cassandra, HBase, Hadoop, Hive, Scribe, BigPipe, Varnish, Haystack and other components together enable handling over 60,000 servers, 300 TB of cached data, 1 trillion daily clicks and petabytes of storage.
An in‑depth interview with Facebook PM David Shein reveals why he left consulting, how he thrived without coding skills, and the key strategies non‑technical professionals can use to excel in product management at tech‑focused companies.
HHVM, Facebook’s high‑performance PHP virtual machine, uses a type‑driven tracelet JIT and custom bytecode to achieve up to nine‑fold speed gains, making runtime optimization the only practical solution for Facebook’s massive codebase while requiring careful compatibility testing, memory management, and Hack language support.
