Uber Case Study - VMware Monitoring

Uber is the largest software platform in the industry connecting drivers with riders, and rated as one of the most disruptive companies in the world. Available in close to 100 countries, with over 100 million monthly average users, Uber’s IT infrastructure has stringent real-time needs. Cameron Soleimani of Uber describes their Sensu deployment.

The Problem

Uber had a large VMware environment with thousands of VMs, along with AWS and looking at Kubernetes in the near future. They were heavily invested in Infrastructure as Code using Puppet, and was looking for a modern cloud monitoring solution that could support multi-cloud and integrate with Puppet and point solutions such as Graphite and Elastic.

Cameron’s team was responsible for monitoring but there were multiple developer groups that need monitoring. They needed a monitoring solution for multiple teamsso they could offer “monitoring as a service”.

Their existing SolarWinds deployment was not devOps friendly and did not meet these requirements.

The Solution

Cameron was looking at Zabbix, but then learnt about Sensu when a co-worker at an earlier job showed it to him and he was impressed with its features. Not only did Sensu integrate well with Puppet, Graphite and Elastic, but it also has a large user community and monitoring plugins collection.

It allowed him to offer “Sensu as a service” to the different developer teams in Uber. The customization and granularity allowed the teams to monitor whatever they wanted and setup rules to send different alerts to different team members.

“Sensu makes me feel like I can sleep well at night”

Cameron Soleimani, Systems Engineer @ Uber