Three Things we Learned at Serverless Conf Austin
Serverless is a hot tech trend and gets a lot of buzz. Serverless Conf investigates the substance.
As previous supporters of, and presenters at, Serverless Conf, last week saw us once again in Austin. This global developer community conference focuses on sharing knowledge and experiences on building applications with serverless architectures.
There was a great roster of sessions this year, including our very own Dragos Dascalita Haut and Ryan Stewart from the Adobe I/O team. Our topic, “Why we made Serverless Deploymentless with OpenWhisk” shared our ongoing journey at Adobe with serverless, how we’re thinking about it at Adobe, and our commitment to OpenWhisk.
Besides encouraging people to contribute to OpenWhisk, we also wanted to learn more about the latest in serverless and what other people were doing with it. Here’s some things we learned:
Defining the Evolution of Serverless
Many presenters talked about the history of serverless or had a definition slide of serverless, because each person had a different take on it and areas to highlight. Since serverless is new, presenters also took the opportunity to talk about tooling and operations. Jason McGee from IBM (above) showed the evolution as: Bare Metal, Virtual Machines, Containers and Functions. As less time is spent setting up the environment, more time is focused on the actual functionality.
It’s Still Early Days for Serverless
Charity Majors got a lot of laughs with her talk “Serverless: The Toddler Years,” which summed up the sentiment at the conference. There’s a lot of work being done in serverless, but there’s still a lot of growing to do. One example: Guy Podjarny highlighted key security concerns for a serverless world. By eliminating infrastructure management, security concerns switch to the platform provider. There are areas where this is better, neutral or worse. Check out Guy’s slides and read more about his views on security.
Exploration over Uses Cases
One of the questions that many attendees had was “How can I use serverless?”. Most people we talked to in the audience were in the exploration and learning phase. They were hungry to learn more about what they could do with serverless and how to do it.
But because serverless is so young, we haven’t yet reached a state where we have lots of users sharing their stories. Some presenters did show reference implementations or example use cases. Given the potential of serverless, we can’t wait to see the growth of user stories at future events.
If serverless is out of its infancy and graduated to toddler status, where do we go from here? Listening to the presenters and talking with attendees, we know that there is a lot of enterprise adoption, more tools and features, and broader integration with services on the horizon. Serverless is a growing ecosystem, and there’s still so much more to come. We’re excited to incorporate serverless into our own platform.