A Study Determining Which Cloud Provider Works Best For Specific Tasks

Those of us who are building web apps and services inevitably ask ourselves “if I build in the cloud, which one performs the best?”.  In this scenario, the word “best” is a misnomer – what we really mean is “what cloud can perform better given the particular needs of my app?”.  Well, Alistair set out to find those answers.

Along with Webmetrics, he published research that quantifies web application performance on Amazon, Google, Salesforce, Rackspace and Terremark.  Instead of using a “one size fits all” methodology, he compared service response, network performance, CPU, and internal I/O for each cloud provider.

Here’s a quick summary of the results; you can download the full study — complete with detailed conclusions, test methodology, and even agent code –for free from Webmetrics.

No study like this has been done before – and I hope that many people will take the code, improve it, and run tests of their own.  Oh, and it’s all free.  Here’s the original post on the topic.

Enjoy :)

Metrics 101 at Velocity

The funny thing about monitoring is that there’s a ton of data to collect, but few people know where to start. A few months ago, Steve Souders, one of the co-chairs of Velocity, asked us if we’d teach a workshop to try and fix this.

It’s a reworking of many of the things we cover in the book, plus an attempt to explain the math and reporting in an accessible way. Here’s the slide deck.