Architecture Weekly Issue #168. Articles, books, and playlists on architecture and related topics. Split by sections, highlighted with complexity: 🤟 means hardcore, 👷♂️ is technically applicable right away, 🍼 - is an introduction to the topic or an overview. Now in telegram and Substack as well.
Business Oriented System Design Course Cohort #5 is closed, but you can apply for Cohort #6(starting 23rd of July) waiting list here.
Highlights
Techniques for improving text-to-SQL 👷♂️
Yeah, this is not AI-related newsletter, but you can't ignore it. Especially if are talking about a blue-print for SQL generation from the text. Context problems, SQL dialects and many more in this brilliant article from Google.
T
#ai #sql
Understanding TLS: how data stays secure over the internet 🍼
Peter Levels recently highlighted how VPN companies want you to think you're naked out there without them; but engineers should actually know how TLS grants you a decent level of encryption protecting your communications with remote server. Study how this happens: that's essential knowledge.
Franco Fernando
#security
Migrating Uber’s Compute Platform to Kubernetes 👷♂️
Uber migrated from Apache Mesos to Kubernetes. Container Platform team at Uber now manages over 50 clusters, each with 5,000–7,500 hosts, supporting 4,000 services and handling 1.5 million pod launches daily. To address scalability challenges, they optimized Kubernetes components, such as tuning QPS settings and switching to Proto encoding. The migration was designed to be transparent to developers, utilizing Uber’s service federation layer, Up, to automate transitions without disrupting workflows.
#kubernetes
Follow-Up
Powering Apache Pinot ingestion with Hoptimator 👷♂️
Traditionally, integrating data into Pinot required manual efforts from data producers to tailor streams for optimal ingestion, often leading to redundant processing and increased operational overhead. With Hoptimator, Pinot transitions to a consumer-driven model, allowing it to autonomously generate and manage its ingestion pipelines based on real-time needs.
Authored byRyanne Dolan
#dataengineering
The types of companies you can work for and what they do for your career 🍼
Not sure, if I worked for all of them, but part of them definitely and I can testify the guide is pretty truthful. Discover what to expect from early startups, old enterprises and scale-ups.
Elena Verna
#work
A Survey of B-Tree Locking Techniques 👷♂️
And can't let you without a paper - this one, written by Goetz Graefe, explains different ways to manage access to B-trees — a common way databases store and organize data. It looks at how to let many users safely read and write data at the same time without causing problems. The paper compares locking techniques, like locking whole sections of the tree or just the spaces between keys. It also talks about special tricks like “ghost records” (temporary placeholders) to make updates faster and safer. Overall, it’s a helpful guide for anyone building high-performance databases.
Authored byRyanne Dolan#db
The cryptography behind passkeys 👷♂️
Passwords are slowly dying thanks to passkeys - a browser technology which gets rid of the very need of password. But how this magic works? Find out the structure, protocols and the mechanisms to fight threads in this insightful post.
Joop van de Pol
#security
Big thanks to Nikita, Constantin, Anatoly, Oleksandr, Dima, Pavel B, Pavel, Robert, Roman, Iyri, Andrey, Lidia, Vladimir, August, Roman, Egor, Roman, Evgeniy, Nadia, Daria, Dzmitry, Mikhail, Nikita, Dmytro, Denis and Mikhail for supporting the newsletter on Patreon!