Architecture Weekly #179
Architecture Weekly Issue #179 . 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.
Highlights
Dynamo, DynamoDB, and Aurora DSQL 👷♂️
Those 3 databases represent 3 generations of distributed highly scalable databases at AWS. It's so interesting how they handle durability, availability and consistency requirements. Grab a great as always post by none other than Marc Brooker.
#db #aws
The anatomy of LSM Tree 🤟
The Database Internals book taught me how B-Trees and B+Trees work. This post maintains the same high level of the quality explaining how LSM tree version data, provide durability guarantees alongside with flash quick write performance.
#db
Top 5 diagrams for documenting software architecture 🍼
I refreshed my old article about documenting software architecture with 5 types of diagrams from C4 context ones to the use case diagrams. Find out why you need them, how to draw them and hear some feedbacks about using those diagrams. Tools recommendations included.
#documentation
Follow-Up
OpenFreeMap survived 100,000 requests per second 👷♂️
I truly wish everyone to experience 100,000 QPS and only see a tiny problem of some static data does not always load. Interesting case study from the OpenStreetMap, how they got that crazy load and how Cloudflare helped to remediate the issue.
#casestudy #performance
Removing friction from Amazon SageMaker AI development 🍼
Interesting note on the challenges that SageMaker team faces trying to provide secure environment for machine learning with the access to giant compute resources - scale, security, observability - everything is here.
#o11y
Scaling Postgres to the next level at OpenAI 👷♂️
A video for a change: OpenAI share how they deal with write amplification, single point of failure and inefficient queries while using a single master setup for ChatGPT. Useful stuff with no talking about AI!
#db
Parquet Content-Defined Chunking 👷♂️
And the last one for today: HuggingFace tells us the story of optimizing their 4 PB storage occupied by Parquet files through smart content-defined chunking as even small changes will trigger significant changes in byte representations.
#dataengineering
Observability for Claude Code
Having AI writing code for you may sound like a productivity boost until you face a limit mid-sprint or get an unexpectedly long bill at the end of the month. The remediation? Proper observability indeed. Grab a piece explaining how to set it up for Claude Code.
#observability
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!