🏆 Most Read Tech Articles In 2023

Most read tech articles across engineering blogs in 2023 featuring Meta, Airbnb, Netflix, Slack and more!

Happy Thursday 👋

As it’s the last Thursday of 2023, I compiled a list of the most read articles across engineering blogs in 2023.

Putting this list together was not an easy task as there’s no simple way to get the number of “reads” of an article. They tend to be published across different blogs powered by various engines.

To tackle this problem, I took into consideration the engagement across Hackernews, Reddit, X as well as the number of opened links in the previous Big Tech Digest newsletter issues. With some help of Python and Jupyter, I’m excited to share the final list!

Before we dive in, I would greatly appreciate if you tweet or mention this newsletter to your connections and follow me on X 🙏 This helps me grow the newsletter while keeping it free.

// 📫 Most read articles in 2023

1. 🥇 "How Meta built the infrastructure for Threads"

by Laine Campbell, Chunqiang (CQ) Tang ⸱ Meta ⸱ 9 min read ⸱ 19 Dec 2023

• Discusses the successful launch of Meta's Threads and the infrastructure behind

• Describes the use of ZippyDB, a distributed key/value database, and how it was optimized for the Threads launch

• Explores the role of Async, a serverless function platform, in scaling workload execution for Threads

2. 🥈 "Slack’s Migration to a Cellular Architecture"

by Cooper Bethea ⸱ Slack ⸱ 9 min read ⸱ 22 Aug 2023

• Tells a story about migration from monolithic to cell-based architecture at Slack

• Introduces the concept of gray failure in distributed systems

• Explains how Availability Zones can be drained

• Covers the implementation of siloing and traffic-shifting in cellular architecture

3. 🥉 "Migrating Netflix to GraphQL Safely"

by Jennifer Shin, Tejas Shikhare, Will Emmanuel ⸱ Netflix ⸱ 8 min read ⸱ 14 Jun 2023

• Describes the migration of Netflix's iOS and Android apps to GraphQL with zero downtime

• Explores the use of three key testing strategies: AB Testing, Replay Testing, and Sticky Canaries, to ensure a safe and smooth migration

• Covers the phased approach to migration, including the creation of a GraphQL Shim Service and the subsequent transition to GraphQL services owned by domain teams

• Discusses the challenges and wins of each testing strategy

• Shares insights into the tools developed, such as the Replay Testing framework and Sticky Canaries, to validate functional correctness, performance, and business metrics during the migration

4. "What is an inverted index, and why should you care?"

by Charlie Custer ⸱ Cockroach Labs ⸱ 7 min read ⸱ 17 Aug 2023

• Describes how inverted indexes work and their impact on database performance

• Explores the downsides of using inverted indexes, specifically the minimal impact on write performance

• Covers how to use inverted indexes, including when and how to create them

• Shares examples and best practices for using inverted indexes in relational databases

5. "Scaling the Instagram Explore recommendations system"

by Vladislav Vorotilov, Ilnur Shugaepov ⸱ Meta ⸱ 11 min read ⸱ 9 Aug 2023

• Discusses the use of Machine Learning in the Explore recommendation system on Instagram

• Describes the use of Two Towers neural networks to make the recommendation system more scalable and flexible

• Explores the use of task-specific DSL and a multi-stage approach to ranking in the system

• Covers the use of caching and pre-computation with Two Towers neural network to build a more flexible and scalable ranking system

• Introduces techniques such as Two Tower NN and user interactions history in the retrieval stage, and the use of Bayesian optimization and offline tuning for parameters tuning.

6. "Understanding Real-Time Application Monitoring"

by Ritesh Kapoor ⸱ Expedia Group ⸱ 7 min read ⸱ 13 Jun 2023

• Covers the performance indicators and SLI/SLO/SLA concepts for application monitoring

• Shares different categories of metrics, including application VM, API, database response, infrastructure, and more

• Explores the importance of monitoring distributed tracing for troubleshooting requests with high latency or errors

• Gives an overview of the challenges of improving operational performance and the benefits of monitoring applications with the right metrics and tools

7. "Improving Performance with HTTP Streaming"

by Victor ⸱ Airbnb ⸱ 7 min read ⸱ 17 May 2023

• Describes how HTTP Streaming can improve page performance and how Airbnb enabled it on an existing codebase

8. "How does B-tree make your queries fast?"

by Mateusz Kuźmik ⸱ Allegro ⸱ 12 min read ⸱ 27 Nov 2023

• Introduces B-Tree as a data structure and clarifies B-Trees vs. BSTs

• Explains B-Tree organization and search queries

• Explores the practical implications of using B-trees on hardware, including CPU caches, RAM, and disk storage

• Explains how packing multiple values into a single node reduces random access and enhances query performance

• Addresses balancing in a B-Tree

9. "Meta developer tools: Working at scale"

by Neil Mitchell ⸱ Meta ⸱ 4 min read ⸱ 27 Jun 2023

• Describes Sapling, an open-source version control system designed for extreme scale

• Covers Buck2, a build system supporting remote caching and execution for large-scale development

• Explores testing and static analysis tools used at Meta, including Infer, RacerD, and Jest

• Presents Sapienz, a tool for automatically testing mobile apps

10. "How Gradle Reduced Build Scan Storage Costs on AWS by 75%"

by Oliver White ⸱ Gradle ⸱ 4 min read ⸱ 23 Jun 2023

• Describes the challenge faced with inefficient cloud storage using Amazon RDS

• Presents the decision to migrate to Amazon S3 as the solution

• Shares the immediate 75% reduction in cloud expenses as a result of the migration

• Explains the added benefit of enabling automatic deletion for unactivated scans after the migration

11. "Real-time Messaging"

by Sameera Thangudu ⸱ Slack ⸱ 7 min read ⸱ 11 Apr 2023

• Describes the architecture used to send real-time messages at scale

• Discusses the setup of the Slack client, including the use of Webapp, Envoy, and GS to establish a websocket connection

• Explains the process of broadcasting a message to all online clients following the journey of the message through the stack

• Covers the different types of events, including regular traffic spikes for reminders, scheduled messages, and calendar events

12. "How Discord Stores Trillions of Messages"

by Bo Ingram ⸱ Discord ⸱ 3 min read ⸱ 6 Mar 2023

• Describes problems with a Cassandra database storing billions of messages

• Covers the impact of hot partitions on latency and end-user experience

• Shares the challenges of cluster maintenance tasks and compactions

• Discusses the frequent tuning of JVM's garbage collector and heap settings to address latency spikes

Thanks for reading 🙏

I would greatly appreciate if you tweet or mention this newsletter to your connections and follow me on X 🙏 This will help me grow the newsletter while keeping it free.

Have a great time during the rest of the festive season!