Architecture Weekly #119 - 20th March 2023

Welcome to the new week!

Postgres is a shapeshifter. In Marten, we're changing it into a document database and event store. We provide built-in read models. Projecting events straight into Postgres gives you enough power for most common cases. Yet, sometimes you need to do more.

I wrote in my last article about how you could build a projection to other databases using Elasticsearch as an example:

• Oskar Dudycz - Projecting Marten events to Elasticsearch

Gunnar Morling also wrote an incredible guide on how to use Postgres Write-Ahead Log as a source of events, notifications and a backbone of the message flow.

• Gunnar Morling - The Wonders of Postgres Logical Decoding Messages

Postgres is great, but do you know what’s the most deployed relational database? It’s SQLite. It can be used, and it’s used anywhere. You can also run it on edge, IoT, and WebAssembly, making it potentially a building block of the decentralised infrastructure. That’s what Philip O'Toole is trying to do with RQLite. Read more in his lessons learned from building it last nine years:

• Philip O'Toole - 9 years of open-source database development

Are databases must haves?

What if I told you you could scale your system to millions of users using Google Sheets as storage? Ok, I won’t. I’ll let Levels.fyi told you how they do it:

• Levels.fyi - How Levels.fyi scaled to millions of users with Google Sheets as a backend

Now homework for you. Could you do it with your system by replacing your database with some other eccentric type of storage? How would you approach that if you had to?

Interesting Architecture composition with unusual building blocks is explained in the following article:

• Zied Ben Tahar - Building an AI powered and Serverless meal planner with OpenAI, AWS Step functions, AWS Lambda and CDK

It’s worth doing mind exercises by putting a limitation on you and thinking about how you could approach your design without having all the basic building blocks that you just blindly assume. Not always will something new show up, but you may be surprised by what you find.

Read also a nice explanation of how ATM infrastructure works. This can also be nice gymnastic as it also might not be how you expect it:

• Patrick McKenzie - The infrastructure behind ATMs

Check also other links!

Cheers
Oskar

p.s. I invite you to join the paid version of Architecture Weekly. It already contains the exclusive Discord channel for subscribers (and my GitHub sponsors), monthly webinars, etc. It is a vibrant space for knowledge sharing. Don’t wait to be a part of it!

p.s.2. Ukraine is still under brutal Russian invasion. A lot of Ukrainian people are hurt, without shelter and need help. You can help in various ways, for instance, directly helping refugees, spreading awareness, and putting pressure on your local government or companies. You can also support Ukraine by donating, e.g. to the Ukraine humanitarian organisation, Ambulances for Ukraine or Red Cross.

---

Architecture

• Levels.fyi - How Levels.fyi scaled to millions of users with Google Sheets as a backend

• Patrick McKenzie - The infrastructure behind ATMs

• Jacopo Romei - EventStorming from the trenches

• Barry O'Reilly - An Introduction to Residuality Theory

• StackOverflow Blog - After the buzz fades: What our data tells us about emerging technology sentiment

• Paolo Banfi - EventSourcing & CQRS: a light introduction

Databases

• Gunnar Morling - The Wonders of Postgres Logical Decoding Messages

• Gwen Shapira - Things DBs Don't Do - But Should

• Philip O'Toole - 9 years of open-source database development

DevOps

• Charity Majors - Deploys Are the WRONG Way to Change User Experience

Testing

• Mike Bland - Test Mercenaries

AI

• Google - A new era for AI and Google Workspace

• InfoQ - Google AI Updates Universal Speech Model to Scale Automatic Speech Recognition beyond 100 Languages

AWS

• Zied Ben Tahar - Building an AI powered and Serverless meal planner with OpenAI, AWS Step functions, AWS Lambda and CDK

• AWS - EKS Workshop - Practical exercises to learn about Amazon Elastic Kubernetes Service

JavaScript

• Jason Godesky - When JavaScript Fails

.NET

• Oskar Dudycz - Projecting Marten events to Elasticsearch

• Stephen Toub - How Async/Await Really Works in C#

• João Antunes - Mapping ASP.NET Core minimal API endpoints with C# source generators

• Steve Gordon - An Introduction to Elasticsearch for .NET Developers

• Steven Giesel - How to write your own cron Job scheduler in ASP.NET Core (like Quartz, Hangfire, ...)

TypeScript

• Microsoft - Announcing TypeScript 5.0

Coding Life

• David Golden - How to mentor software engineers

Management

• Aino Corry - Retrospectives Antipatterns

• Trond Hjorteland - Sociotechnical Systems Design for the “Digital Coal Mines”

• Harvard Business Review - How Chinese Companies Are Reinventing Management

Industry

• Gergely Orosz - The Collapse of Silicon Valley Bank

• Gergely Orosz - Amazon doubling down on RTO (return to office)

• OnVoard - Our production servers was suspended by Google Cloud