Marc Brooker's Blog

Better Benchmarks Through Graphs Isn't the ambiguity in the word *graphs* fun?

How Do You Spend Your Time?

Pat’s Big Deal, and Transaction Coordination Working together towards a common goal.

What is Scalability Anyway?

Why Aren’t We SIEVE-ing?

It’s About Time!

Optimism vs Pessimism in Distributed Systems What—Me Worry?

Writing For Somebody Who's there?

Exponential Value at Linear Cost What a deal!

On The Acoustics of Cocktail Parties Only parties of well-mannered guests will be considered.

Invariants: A Better Debugger?

My Favorite Bits of OSDI/ATC’23 Talking to 3D people is cool again.

Bélády’s Anomaly Doesn’t Happen Often Anomaly is a really fun word.

What is a container?

Container Loading in AWS Lambda Slap shot?

Open and Closed, Omission and Collapse Were you born in a cave?

The Four Hobbies, and Apparent Expertise Around the end of high school, I started to get really into photography.

Surprising Scalability of Multitenancy When most folks talk about the economics of cloud systems, their focus is on automatically scaling for long-term seasonality: changes on the order of days (fewer people buy things at night), weeks (fewer people visit…

False Sharing versus Perfect Placement This is part 3 of an informal series on database scalability.

Hot Keys, Scalability, and the Zipf Distribution the: so hot right now.

NoSQL: The Baby and the Bathwater Is this a database?

Erasure Coding versus Tail Latency There are zero FEC puns in this post, against my better judgement.

Under My Thumb: Insight Behind the Rules My left thumb is exactly 25.4mm wide.

Lambda Snapstart, and snapshots as a tool for system builders Clones.

Amazon’s Distributed Computing Manifesto Manifesto made manifest.

Writing Is Magic Magic can be dangerous.

Give Your Tail a Nudge Tricks are fun.

Atomic Commitment: The Unscalability Protocol 2PC is my enemy.

Histogram vs eCDF Accumulation is a fun word.

What is Backoff For?

Getting into formal specification, and getting my team into it too Getting started is the hard part Sometimes I write long email replies to people at work asking me questions.

The DynamoDB paper The other database called Dynamo This week at USENIX ATC’22, a group of my colleagues1 from the AWS DynamoDB team are going to be presenting their paper Amazon DynamoDB: A Scalable, Predictably Performant, and Fully Managed NoSQL Databa…

Formal Methods Only Solve Half My Problems At most half my problems.

What is a simple system?

Simple Simulations for System Builders Even the most basic numerical methods can lead to surprising insights.

Fixing retries with token buckets and circuit breakers Throttle yourself before you DoS yourself.

Will circuit breakers solve my problems?

Software Deployment, Speed, and Safety There's one right answer that applies in all situations, as always.

DynamoDB’s Best Feature: Predictability Happy birthday!

The Bug in Paxos Made Simple There's not really a bug in Paxos, but clickbait is fun.

Serial, Parallel, and Quorum Latencies Why are they letting me write Javascript?

Caches, Modes, and Unstable Systems Best practices are seldom the best.

My Proposal for Arecibo: Drones With apologies to real radio astronomers Last night I finally got around to watching Grady Hillhouse’s excellent video on the collapse of the Arecibo Telescope.

Latency Sneaks Up On You And is a bad way to measure efficiency.

Metastability and Distributed Systems What if computer science had different parents?

Tail Latency Might Matter More Than You Think A frustratingly qualitative approach.

Redundant against what?

What You Can Learn From Old Hard Drive Adverts The single most important trend in systems.

Incident Response Isn’t Enough Single points of failure become invisible.

The Fundamental Mechanism of Scaling It's not Paxos, unfortunately.