“What is Serverless?” Presentation

A few weeks ago I was invited to speak about serverless computing at two meet-ups. I have been honing a What is Serverless presentation that I gave for the first time. I want to educate people on what serverless is and where it makes sense.  Finally, I want to warn them about where the gotchas might be for adopting a function-as-a-service environment. I was graciously hosted by Todd Lewis the producer of All Things Open and Great Wide Open. (By the way, the call for papers for All Things Open just opened up. It’s an amazingly well-run open source conference in my home town of Raleigh, NC in October that I highly recommend.)

I spoke at All Things Open RTP Meetup. This was my first shot at giving a serverless talk. Thanks to some feedback I updated my presentation for the Open Source South Carolina.

Here’s the abstract:

Serverless can be misleading as a descriptor. Actually, as a descriptor, it’s almost divisive.  Serverless infrastructure actually runs on servers. However, the “server-less” reference comes from the fact that serverless abstracts the complexity of running servers away from the software developer which enables them to develop software without having to worry about the scaling, redundancy and overall infrastructure design. This is called Function-as-a-Service or FaaS.
For the purposes of this talk, we’ll discuss serverless technologies where someone else is providing serverless infrastructure. Popular serverless platforms include Amazon Web Services Lambda, Google Cloud Functions, and Microsoft Azure Functions.
The presentation will also discuss the software that can be used to deliver Functions-as-a-Service (FaaS) that enables serverless, including serverless frameworks like Knative, Kubeless, OpenFaaS, and Oracle’s fn.
Finally, we’ll cover what a cloud-native application might look like including the use cases and design patterns that serverless is geared towards providing.

This presentation is a work in progress. I got a lot of great questions that I am working to include. One of those areas is observability. I will be working on updates on the challenges faced by monitoring and tracing ephemeral infrastructure. In addition, I received quite a few questions about stateful vs. stateless functions. I think that is one thing that many people are having a hard time grasping, I will work on updates to the presentation for my next talk at the Triangle Kubernetes and OpenShift Meet-up on March 21st.