Predicting Bluesky’s Scale with Jaz

02/14/25 • 65 min

Bluesky has been on a roller coaster of growth for over a year. From the early days of figuring out a new distributed social protocol—AT protocol—to actually building it and inviting 30 million of their closest friends. Not only has the site gone through tremendous growth, the team has been optimizing, re-architecting, and adding features the entire time.

Jaz is a software engineer focused on the infrastructure at Bluesky, and they share how they achieved exponential growth without exponential costs. We cover some of the key components of the protocol and how that affects the architecture.

There’s some amazing advice from the trenches we know you’ll enjoy.

Show Highlights
(0:00) Intro
(5:00) Jaz’s background
(12:30) Bluesky Infrastructure
(17:00) Predicting the future
(20:00) What is a PDS?
(22:30) Relay and firehose
(26:00) Work queues
(30:00) Scaling physical servers
(37:00) How do you handle incidents?
(41:00) Where’s Kubernetes?
(43:30) How video changes
(45:00) Data locality
(46:30) Hardware decisions
(53:00) What bad decisions?
(57:00) Launching video
(1:00:00) What’s next?

About Jaz

Jaz is a software engineer who learned from on-the-job experience. They have a background with hardware which makes them better with software. If they’re not drinking Monster they’re building a single purpose database, or maybe they’re doing both. Jaz went from building with AT protocol to building AT protocol in a matter of months. They also have an impressive collection of plushies and power tools.

Sponsor the FAFO Podcast!

http://fafo.fm/sponsor

There’s some amazing advice from the trenches we know you’ll enjoy.

About Jaz

Sponsor the FAFO Podcast!

http://fafo.fm/sponsor

Previous Episode

Making Smart Infrastructure Decisions with Lauren Long

Today we find out how building a product at Big Tech can be very different than a startup. Lauren Long has done both. Building parts of Firebase and eventually taking the things she learned to build Ampersand. We discuss what Ampersand is and go into detail about what the back end looks like. We even drop some hot takes about serverless and Kubernetes. We think you’ll love it!

Show Highlights

(0:00) Intro

(1:00) What is Ampersand?

(3:00) What is the backend?

(4:00) What is Lauren’s background

(6:00) How are people using it?

(10:00) How is Temporal used?

(14:00) How to keep APIs in check

(21:00) What did you learn?

(24:00) What has broken?

(26:00) Why use Kubernetes?

(32:00) What have customers done?

(38:00) What’s next?

About Lauren Long

Lauren Long is the CTO and co-founder of Ampersand, an API integration and workflow engine for enterprises to integrate their data with hundreds of applications. Lauren co-founded Ampersand after working at Google on their serverless products and saw a need for a different kind of integration for customers. She’s a developer with a great intuition on how to build reliable and scalable systems.

Sponsor the FAFO Podcast!

http://fafo.fm/sponsor

Next Episode

Animating the Stack with Sam Rose

This episode is stacked with information. You could even say “full stacked.” Sam has built and run some large scale systems as a SRE at Google, now building backend services at Budibase, and he spends his free time teaching others how systems work at understandable scale. We dive into what makes Google SRE different from other companies, what it’s like to be a parent, and how Sam got started with building animations for his blog. Don’t forget to visit and check out the easter eggs he’s hidden throughout.

Show Highlights

0:00 - Intro
2:00 - Sam’s background
6:00 - How Google did SRE
15:00 - Importance of docs
19:00 - The problems with Java
26:00 - Budibase
32:00 - Borg vs Kubernetes
39:00 - Building animations
46:00 - Being a better teacher
56:00 - Art in the age of AI
1:00:00 - What’s next

Links Referenced