New Post: Estimating Availability - Four 9s is the Ceiling
AKF Partners Logo Technology ConsultingScalability - We wrote the book on it ℠


Short videos that make a difference

AKF Scalability Cube - explained

LinkedIn Post · Subscribe on
YouTube Logo YouTube

The Scalability Cube Explained

In the books, The Art of Scalability and Scalability Rules, the founder and CEO of AKF Partners, Marty Abbott, outlines key points to ensure your applications are scalable today, and in the future with growth.

The Scale Cube is a 3-dimensional representation of the three basic approaches to scalability. We have the X-axis, the Y-axis and the Z-axis.

The first strategy, or x-axis, is the most common way to split a monolithic application. It involves running multiple copies of an application or service behind a load balancer to distribute those requests for greater capacity.

Think of it as a checkout line in a popular department store. If you had a lot of customers but one checkout counter, that line could be long and slow.

However, opening up additional checkout counters and splitting the customers to the various lines, reduces the demand on the first checkout counter, and speeds up the line for everyone else.

The second strategy, the Y-axis, is about dividing tasks based on what you know about either the service or data involved. For your application, this might mean breaking up the services by functional area like account creation, checking out, or order status.

And in our department store, this could mean breaking out the various counters into certain ones for checking out, others for returns, and maybe another one for picking up online orders, etc. Still the same number of customers in the store, just breaking them up to functional areas to reduce bottlenecks.

The third strategy, or the Z-axis, is all about separating elements according to customer or object knowledge. This could be breaking up customer requests based on the country they live in, or their time zone, or maybe something as simple as their user ID.

And if we bring this back to our department store, this would be like splitting your customers into different physical stores based on where they live. And just like in the real-world examples of a department store, within each store you would still break your customers up by their functional area, checking out, returns, picking up orders and also, by creating more than one checkout counter to reduce bottlenecks.

For a deeper dive on the topic of the Scale Cube, or scaling your applications for growth, pick up a copy of our books “The Art of Scalability” or “Scalability Rules” or feel free to reach out to us at We are here to help.

Next Edition: Technical Debt - Part I


Most Popular: