Mesosphere DC/OS

Role: While at Mesosphere I lead the majority of my design projects. Based on sheer size from time to time, some were supported by one product designer and guided (if necessary) by our head of design. We swept across all areas of the product as I lead our design system efforts along the way.

 

Mesosphere DC/OS Installation in process

Mesosphere DC/OS installation success

 

Services

DC/OS at its core runs a set of services distinct to every companies datacenter. We present a simplified and complete summary of those services to provide details about up-time, resources, and activity. In the event of extremely complex installations, a thorough filtering mechanism was thoughtfully created.

maintenance and operations

Our primary persona was a DevOps expert. It was exceedingly important for our users to maintain healthy statuses. Integral in this effort was having the ability to troubleshoot instantly. We went to great lengths to provide details changelogs, activity logs, configuration difs, and the ability to revert any change.

 

Why Redesign?

As customer’s needs grew, we quickly realized our product need to grow with it. Some problems we addressed in our redesign included:

  • Scalable Navigation

  • Optimizing real estate for the primary information

  • Reduce the number of bars are establish a stronger hierarchy

  • Update our design system and establish master file library.

 

So… what exactly is a datacenter operating system?

I’m so glad you asked! Here’s a handy dandy video. In our own words, Mesosphere DC/OS automates rollout and production operations for containers and data services.

 
 

Mesosphere DC/OS Roundtable

I worked on an extraordinary amount of initiatives at Mesosphere. Rather than breakdown each project, here’s some highlights and a few extra details worth calling out.

 

Service metrics

The DC/OS Dashboard provides metrics across the entire system. However, when diagnosing issues the details metrics of a service can prove extremely valuable. Here, customers can monitor spikes and understand their resource allocation and usage at varying intervals.

audit logs

We provide output and error logs across all services. Creating a text editor is a difficult problem, and developers tend to have their favorites. So, we provided the simplest output allowing users to download and explore where they saw fit.

Service catalog

The service catalog allowed users to install packages or create their own containers on their system. It’s akin to the iPhone App Store except way more technical.

See Project