5 min read

[P&R#2] KubeCon (NA) 2020 Highlights

Quick shoutout to The Linux Foundation for the amazing work they've done! I would like to express my heartfelt gratitude for being awarded a scholarship to attend the event.

Day 3

I'd like to dedicate my day three highlights to talks from the community segment. We must always remember that at the end of the day, what matters isn't the code; it's the humans involved. That includes yourself! So take care of your mental health, stay safe, and happy coding.

Thoughts: Always remember to take breaks! Also, donating projects to the CNCF is mainly for organizations who need a neutral platform to host their projects. The value of open source can be broken down into three categories: strategic benefits, operation benefits and second-order benefits. And often times, the detriments of open source are smaller than what we make them out to be.

Other notable talks include:

How to Effectively Manage Kubernetes in a Regulated Environment - CapitalOne

  • Critical Stack: Allows developers to focus on writing code and not managing infrastructure, and eliminates out-of-the-box configuration challenges during Kubernetes set-up

Inside Kubernetes Ingress - Dominik Tornow, Cisco

Includes two types:

  • Network ingress: How to get traffic into the cluster
  • Kubernetes ingress: How to route traffic within the cluster

Day 2

Whelps, I'm sorry it took awhile to update my highlights from Day 2.

To jump right into it, the biggest highlight for me today was the talk: A Walk Through the Kubernetes UI Landscape. It's a really great talk, and the speakers gave a comprehensive overview of what user interfaces are available for Kubernetes. However, many of these products are in fact very similar to what I'm building with Kubeinn - perhaps even better. These include Octant by VMware, Lens by LensApp, and K9s by Derailed. While it's great to see so many applications like these, I had to take a few days off writing to think about my product once again, and how it fits into our use case.

Another interesting talk which I attended was Stop Writing Operators - Joe Thompson. He shares about why as developers, we should strive to avoid writing operators unless it's absolutely necessary, as operators need to be maintained. Yet this lesson extends beyond operators; in fact, we as developers should strive to reduce the amount of code we write as much as possible. It has also given me a deeper appreciation of no-code and low-code tools. I love this post by Kelsey Hightower - nocode. In addition, when we're thinking about automation, it's wise to think about this diagram here.

xkcd: Is It Worth the Time?

If our automation tool doesn't shave off as much time as we'd like, or if it doesn't deliver other additional benefits, it might not be worth building the automation tool. We'll be saving even more time by maintaining the status quo!

Finally, I'd like to wrap with the talk: More Power, Less Pain: Building an Internal Platform with CNCF Tools - David Sudia, Senior DevOps Engineer, GoSpotCheck, Inc.. The main message here is that the Kubernetes ecosystem has matured so much over the past few years that if you're a team that's thinking of building an internal tool, it's perfectly okay to pick a leading open source tool now that the market has more or less decided on its settlers.

Day 1

I started off the day by attending the TechDocs meeting, which was a small group interactive session with technical writers from the Kubernetes project. Who would have thought that a project could grow so large that it requires a whole team of technical writers to write documentation! It's also one of the first times I've paused to think about the two types of documentation: user and developer documentation.

From the session, I'm reminded once again of the importance of talking about the product even before it's built. It's similar to what I shared in my previous post with Tara Reed. When working on a project, it's good to just go ahead and start a documentation website right off the bat; it helps to narrow the scope for the project.

As an aspiring open source documentation contributor, there are several things we could do. First, try to do everything in public, even when you think you could have done it in private! This opens up the conversation to the rest of the community and allows you to hear their opinions before investing your time into doing the actual work. In addition, even though we might have our own opinions on how the documentation should go, it's worth taking a step back and deferring to the key stakeholders' opinions at times. As the saying goes, have "strong opinions, weakly held". Another tip would to focus on the areas which provide the maximum benefit. And that's so important, because we have to recognize that busyness does not translate into maximum benefit. Finally, always, always, always be kind! We're here to build a community, so let's make it one we all enjoy being a part of.

I attended one breakout session today, which was Contain Your Enthusiasm for Go Dev: Fast, Simple Go+Docker Development - Ed Warnicke, Cisco. It was pretty cool! The speaker shared tips on how we can speed up the build process for Golang application development, as well as how we can introduce tests and debugging development process. I'll be applying these tips to Kubeinn some time in the near future.

I also attended a couple of keynote sessions, but I'll talk more about those as part of my Kubeinn outreach content. Onwards to day 2!

Day 0

KubeCon's finally here! Over the next few days, I'll be sharing some of my highlights from the conference. So far, I've really been enjoying the breakout talks and the maintainers' series. The sponsored lightning talks were pretty cool too, although I'm sure I would have enjoyed them more if they were slightly less pressed for time.

I found a couple of projects that are related to what I'm building with KubeInn. In the Sponsored Lightning Talk: Why You Need Observability to Adopt Kubernetes at Scale - Shreyans Parekh, AppDynamics, the speaker introduced the AppDynamics Cluster Agent, which provides an easy, streamlined approach to collecting cluster-level monitoring data.

Under VMware's sponsor booth, I found an interesting project demo on Pinniped, which provides identity services to Kubernetes. From the way it works, I suppose KubeInn can be thought of as an Identity Provider (IdP) too, since we both aim to enable access to clusters for authenticated users.

Throughout the day, I kept hearing the two words: "vendor-agnostic". It seems that even though there are several companies that are doing managed Kubernetes really well, such as Google Kubernetes Engine (GKE) and Amazon EKS, there's still interest in applications that do not restrict their customers to any particular vendor. It's definitely  something worth thinking about when designing a Kubernetes product.

It was also wonderful to speak with Priyanka Sharma, General Manager of the Cloud Native Computing Foundation, as well as other folks during the AMA Happy Hour. From our conversations, it's so nice to see how warm and welcoming the Kubernetes community is. And I'm really happy to be part of this amazing community.

That's it for today. Stay tuned for more highlights in the coming days!