Building Applications for Operability

Abstract:

We can all agree that breaking down the walls between developers and operations is a fine thing, and doing so is a necessary step along the path to DevOps nirvana. The challenge I've encountered over and over again, though, is that we are coming from very different places – the operations guys often struggle with the basics of software development – testing and version control, for example – and the developers don't really understand what it means to run software in distributed data centers, at scale, when things like to break at the most inconvenient times.

This presentation is an overview of what makes running software in the datacenter different from when a developer runs it on their development boxes, and the things that they can do to help their operations team deploy their applications and keep things humming along. I introduce the potential complexity of an application deployment through an example real-world SaaS service, then work through a checklist of service management interfaces that need to exist for "monitoring" (and why "monitoring" is in quotes).

Speaker:

Reinhardt Quelle

Cloud Services Architect

Cisco