Cisco Meraki sells 100% cloud-managed networking devices and IT solutions that "just work". Meraki products include wireless access points, network switches, security appliances, video surveillance cameras, and phones, as well as software for managing laptops, smart phones, and tablet devices. The more than 3 million network devices in the Meraki cloud all connect to a centrally managed web service called "dashboard" to upload logs and statistics and download configuration. The SREs who work on back-end infrastructure are responsible for managing the more than 800 servers spread across 7 datacentres on 3 continents that are used to run dashboard. In this talk I will give an overview of some of the technical challenges we face and solutions implemented for managing this infrastructure, including spreading our workload, distributing and verifying machine configuration, and collecting and analyzing system logs.

Dr Mark Stillwell is a Site Reliability Engineer working on the Back-end Infrastructure team at Cisco Meraki's London office. He is interested in the problems of resource allocation, scheduling, and modeling for distributed systems. Dr Stillwell was formerly a researcher in the areas of distributed systems and cloud computing at Imperial College London (UK) and ENS Lyon (France), as well as a Lecturer in Distributed Computing at Cranfield University (UK). Mark will be joined by his colleague Patrick Coleman who leads the Back-end Infrastructure team at Cisco Meraki's London office, and is responsible for Meraki's backend data collection and monitoring platforms. Patrick has previously held positions at Deutsche Bank and the Government of Western Australia, working on projects including network monitoring at scale and rapid-response network infrastructure for emergency services working in extremely remote areas.