Distributed Teams
< 5 minute read

Calm Before the Storm

The company behind the world's leading meditation app found itself in a decidedly non-calming situation when an outage had catastrophic consequences.

Tasha Huebner
Tasha is a communications, marketing, and strategy expert working with companies to create insightful and thought-provoking messaging to help them grow their businesses.

Companies such as Calm often implement their own IT solutions, using systems that include the open-sourced version of Kubernetes to deploy, maintain, and scale applications. In spite of issues that included version control and difficulty in sharing across teams, this system worked well enough for Calm until one day it didn’t: a control plane corruption led to Kubernetes failure and a rolling back to the legacy infrastructure.

Before the system failure, Calm had asked Toptal for assistance to manage their existing systems, as they were short on the resources needed to keep up with their exponential growth. Thus, Christopher Stobie, a senior engineer from Toptal’s AWS DevOps practice, was already in place when the system came crashing down on his 2nd day on the job.

Getting to work immediately with Mark Marcantano, Calm’s Technical Program Manager, Mark and Chris were able to build a new fully-automated cluster that would be managed by Amazon Web Services (AWS) instead of self-managed, running under AWS’ new Elastic Kubernetes Services (EKS). Even though he essentially needed to build an entire environment from scratch, Chris collaborated with Mark to take advantage of their existing capabilities; resulting in a migration that was much quicker than expected, in part due to the ease of use of the AWS solution.

The migration to EKS had immediate results, with improved stability and rapid cluster deployment. In layman’s terms, this meant that Calm could turn from focusing on the IT that manages their wellness platform behind the scenes, to instead refocus on the business of signing up new users looking for the meditation and wellness that Calm is known for.

In addition to seamless updates and reliable networking, Calm has benefited from eliminating the outages that existed under the old system, which were costing $40K for each hour the system was down and unable to subscribe users. By finding the exact right person for the job, Toptal was able to partner with Calm and hit the ground running in fixing a serious issue that could have seriously impacted the business had it dragged out longer.

In the six months since the outage, Calm has seen a significant amount of growth, but with its newly implemented system has been able to absorb that growth without any performance issues. While Calm’s guided meditation app is famous for its ability to help people unplug, the new technology behind the app will do anything but, keeping users plugged in for the duration.