Reviving the Edge: ‘Livin’ On the Edge’ is Back

Flynn from Buoyant & Cloud-Native Happenings

In the ever-evolving landscape of cloud-native technology, APIs, Kubernetes, and service meshes, it's crucial to stay up-to-date with the latest developments and insights from experts in the field. That’s why, there’s no better time to revitalize our popular podcast, Livin’ on the Edge. We’re proud to say we’re back and better than ever!

To kick off Season Three, we figured who better to bring on the show than Flynn, the Tech Evangelist from Buoyant. Flynn was one of the original creators of Emissary and a previous Ambassador Labs team member himself. As someone who’s been a part of the Emissary and Edge Stack story from the early days, his 40 year’s worth of knowledge and experience is something we can all learn from.

Flynn’s day-to-day life revolves around dealing with the complexity of the cloud-native world and the rapid pace of change that comes with it. I sat down with him to discuss some of those complexities, as well as what he’s looking forward to right now in the cloud-native world.

Cloud-Native Deployments

One of the complexities he’s dealing with right now is around cloud-native deployments. In the scenario we discussed how CockroachDB, Linkerd, and Emissary are deployed in a multi-cluster setup, reflecting a real-world use case that goes beyond the simplicity of toy setups.

In this setup, Flynn walked me through the intricacies of multi-cluster applications, including data sharing and user redirection based on regions. The complexity of ensuring that multi-cluster applications perform efficiently while adhering to compliance standards is a critical challenge for the cloud-native community.

Speaking of Linkerd and Emissary, we couldn’t have Flynn on the show without highlighting the seamless integration of Linkerd and Emissary and their ability to work together without intricate configurations in order to simplify cross-cluster communication. For those who don’t know, this integration provides an excellent example of how service mesh components can simplify the process for developers and make it easier to navigate the complexities of multi-cluster setups.

2. APIs as User Interfaces

“The idea that your APIs form an interface that humans have to use and therefore they are a user interface is a thing that does not seem to get a lot of attention yet,” shared Flynn.

The next idea that Flynn expects we’ll see more and more of in the coming months is this concept of APIs as user interfaces is a viewpoint that is often overlooked in discussions surrounding cloud-native technology. Typically, APIs are primarily viewed as connectors between applications. However, I think Flynn brings up a great point that many of us forget and he sheds light on the fact that APIs also act as an interface for human interaction.

This is an important note for developers because they must not only understand their products' inner workings but also need to cater to users who lack the technical expertise to do the same. Adding this “APIs as User Interfaces” perspective introduces a new layer of complexity in the development process, making it crucial to create APIs that are not only functional but also user-friendly.

“Every developer should have the experience of having to support an end user who has no expertise in how your product was built,” shares Flynn. “All they know about is what they want to do, and they know your product and see its documentation. Every developer should have to support a nontechnical counterpart trying to get something done with their product because it is a truly eye-opening experience for making you a better developer."

In the evolving landscape of cloud-native technology, it's essential to prioritize user-friendly solutions to make technology more accessible. Focusing on the developers who use these technologies to accomplish their tasks, is a step toward creating more intuitive and user-centric cloud-native solutions.

3. What’s Coming: The Gateway API

In this episode, we also delved into the limitations of the upcoming Gateway API (whose general availability will be launching soon), particularly in terms of handling fundamental features such as retries, WebSockets, and policies. Flynn raised a concern about the use of policy attachment and its complexity, highlighting that less experienced users may find it challenging to configure effectively. But, we shall have to see when GA finally hits in a few weeks!

Honestly, I think the limitations of the Gateway API are due in part to the challenges posed by Kubernetes core changes. When we make changes to the Kubernetes core, it’s a lengthy process, often taking years to implement. That kind of drawn-out timeline can hinder the adoption of new features and functionalities.

Making changes to Kubernetes core also comes with the challenges of slow implementation of core changes. When that’s combined with the need to maintain backward compatibility, it can add an additional layer of complexity. Flynn pointed out that we need a more user-friendly process to implement crucial cloud-native features.

“I think Gateway API should be usable by everybody in lots of different fields,” shares Flynn,

“We should really be thinking about who the Gateway API is designed for and our focus should be on developers and application users.”

What’s Next: KubeCon 2023

Overall, it’s clear the cloud-native community is working diligently to balance the demands of infrastructure experts and developers, creating a space where cloud-native technology serves all of its users effectively. I had a great time chatting with Flynn–who feels like an old friend at this point– over the complex world of cloud-native technology, APIs, and service meshes.

There are a lot of very exciting things that lie ahead for the cloud-native community, even beyond the Gateway API. Many of those exciting revelations will happen at KubeCon 2023, which both Flynn and myself will be at next week!

You can find the Ambassador Labs team at Booth O12, and check out our talk with Flynn on, “Emissary-Ingress: Self-Service APIs and the Kubernetes Gateway API,” happening Tuesday afternoon of the conference from 4:30 pm - 5:05 pm. Cheers to many more Livin’ on the Edge episodes to come, and I’ll see you all at KubeCon!