
How to Make Things Faster is the latest book by Cary Millsap. It’s a book for anybody who is curious about how computer programs and other processes spend their time, and what you can do to improve them.
I can’t believe I was in my forties the first time I saw how to optimize a system the way Cary and his team do it. Now, it doesn’t even make sense to me that anyone would try it any other way.
Richard Russell, former 7-Eleven enterprise architect
Cary Millsap has a gift. Yes, he’s brilliant at making things run faster, but his true genius is translating complex problems into simple, powerful ideas. This master teacher combines vivid stories with clear practices to teach principles that are as timeless and they are timely. If you want systems that keep pace with the new world of work, don’t wait—read How to Make Things Faster!
Liz Wiseman, New York Times bestselling author of Multipliers and Impact Player
Every technology executive responsible for mission-critical systems should read How to Make Things Faster. It’s one of the easiest and most entertaining reads you’ll find, and it will prepare you thoroughly for handling—that is, solving and avoiding—technology performance crises.
Mark Sweeney, former technology executive at Bank of America
Faster is the single most concise material on the topic of performance analysis and optimization out there.
Jonah H. Harris, director of AI & ML at The Meet Group
You may think at first glance that How to Make Things Faster is a book about technology. It’s not. While technology plays an important part, it’s really a supporting player in the story of how people are impacted by technology. Looking at challenges from the perspective of end users is what makes this book so insightful. Cary reminds us that in order to solve real performance challenges, we need to work on what’s most important to the people who interact with the systems we design, build and operate. You’ll meet these people in How to Make Things Faster; and you’ll come away with a new appreciation for how making systems Faster makes things better for all of us.
Dominic Delmolino, vice president at Amazon Web Services
I am fascinated with the way that Cary, through his experiences, tells step by step the path taken to solve performance problems. The essence of How to Make Things Faster comes down to a single word: “Method.”
João Henrique Brogna, database administrator at Pro4tuning Brazil
Contents
- Look at It
- Bob
- Phyllis
- The Real Goal
- Nancy
- Looking at the Right It
- When you Can’t Look at It
- Method
- Forty-Nine Grievances
- The Shipping Labels Problem
- More Grievances
- Priority
- But My Whole System Is Slow
- Collateral Benefit
- The Silver Bullet
- The List
- Method R
- Profiling
- Payroll
- The Sequence Diagram
- The Gantt Chart
- Tracing
- The Profile
- Creating a Profile
- Measuring Performance
- Performance Is a Feature
- The Reproducible Test Case
- Intermittent Problems
- How Much to Trace
- Identifying Experiences
- Measurement Intrusion
- Optimizing
- A Riddle
- A Game
- Event Count
- Event Duration
- Filter Early
- Look Left
- Tow–Millsap Law
- The Bottleneck
- Beware the “System Bottleneck”
- The Problem with Optimizing Subsystems
- Every Problem Is a Skew Problem
- Critical Path
- Delays
- Kevin
- Queueing Delay
- Queueing Theory
- The Hyperbola
- Traffic Intensity
- Utilization
- Hyperbolic Leverage
- Coherency Delay
- The Impact of Delays on Throughput
- Waste
- Debra
- The Ramp
- Martha
- Efficiency
- Fix It, or Feed It?
- Yeti
- Fast vs. Efficient
- Scalability
- Problem Solving
- Four Simple Questions
- Reaching the End of Your Data
- Your C-Level Feedback Loop
- Collateral Damage
- Too Much of a Good Thing
- Doug
- When Are You Done?
- Predicting
- Richard
- Why Predict?
- Predicting with Profiles
- Go/No-Go Predictions
- Linear Behavior
- Skew
- Event Interdependencies
- Nonlinear Behavior
- Latency Hiding
- Mom
- Dominic
- Parallelization
- How to Melt a System
- Multitasking
- Human Multitasking
- Fallacies
- The Evil Genie
- The Leather Jacket
- The Buried Outlier
- Be Careful What You Wish For
- Percentile Specifications
- The Hit Rate Problem
- The MPG Problem
- The Ratio Trick
- Even Throughput and Response Time?
- Are Ratios Useless?
- When Ratios Are Reliable
- Describing Performance Improvements
- The “n Times Faster” Myth
- Testing
- Why Test?
- Risk
- Destructive Testing
- Testing Is Not a Phase
- Automated Testing
- Preventing Problems
- Planning
- Mrs. Utley
- Capacity Planning
- Utilization Targets
- When to Upgrade
- Politics
- Proving
- The Problem with Under-Promising
- Seven Project Risk Magnifiers
- Fail Fast
- Face
- The Jeweler’s Method
- Change Control
- Record Keeping
- Failure
- Don’t Worry, Be Nervous
- Just for Fun
- Optimizing for Kids