Minimum Viable Ops

A97a75c945507f70992f579a730b0657?s=47 Doug Barth
February 24, 2015

Minimum Viable Ops

Building a new system involves a lot of infrastructure and monitoring decisions: where to host, how to deploy, and how to monitor your new application. There's a strong tendency to "do it right" from the get go. But trying to build the ultimate platform will slow down your progress in those crucial first days.

In this talk, we will discuss how to best approach all these decisions. We will explain the various layers of infrastructure you need to successfully run a system in production. These layers include logging, deployments, and monitoring. For each layer, we will discuss options for satifying the needs of those layers, and how those choices will change as your business grows. In the end, I hope to have given you a plan for how to build the first version of your operations platform and some ideas for how that platform will grow and change as your service grows and changes.

This talk was given at the AWS Popup Loft in San Francisco. Unfortunately, the talk was not recorded.


Doug Barth

February 24, 2015