Lessons learned building cloud services using AWS, elasticsearch and MongoDB

About this Presentation

The design of a cloud service is fascinating. Aspects from all directions converge: from billing model to scale to tenancy, everything is important and must be evaluated in context. The reason is that at scale, small problems become big problems. At scale, logging turns into a big data problem, restarting a cluster can turn into hours of re-indexing, the price of a dependency can cripple your billing model.

This session reviews various places in a system where small problems become big problems. We’ll use real examples from CloudHub, an integration PaaS supporting the premier salesforce tool: dataloader.io.

We’ll use CloudHub architecture to tell the story of design for scale, with a focus on rationale and context. You’ll learn about what we designed for, and how we change in response to empirical evidence and customer requirements. Yes, you will see cool tech from folks like AWS, MongoDB and elasticsearch. You’ll also leave with lessons CloudHub learned gluing them together over time.