Yes, we've been there — again! Just like last year, some of us took a couple of days off in order to attend KubeCon 2019.
Still, the event continues to grow year after year and has almost doubled its amount of visitors as compared to last year (is this Moores Law?). 7700 attendees on an open source conference surely is a statement. It underlines the growing importance of container technology, clouds and DevOps in nowadays industry.
The event is a great chance to pick up the latest technology updates on Kubernetes and its ecosystem. It gives new inspiration and connects all the great people of the community. Obviously, this blog post is much too short to cover all the talks in their entirety. Hence, here are some of our personal highlights:
- For years I am admiring Rancher for pumping out innovative tools like K3s (a really small Kubernetes) or RKE (a Kubernetes installer). In this year's talk they presented Longhorn, a distributed block storage solution. In comparison to competitors like Rook (which is primarily known for managing Ceph), Longhorn was developed cloud-native from the start, everything just extends the Kubernetes API by using CRDs. With only 30000 lines of code, it is cut down to the essentials of storage management. And we know that less code is always better. Definitely worth a try!
- The crowd started cheering unexpectedly when the removal of Tiller was announced for Helm 3. The "Kubernetes packet manager" is undergoing heavy changes and refactoring — and I would say it was about time! You should not forget that Helm is around for quite a while. You could say that back then, Kubernetes was still in its infancy. There were no roles nor were there service accounts. Helm is now not only letting go of the legacies of the past but also adds a Lua scripting engine to it's core in order to make Helm Charts more expressive and powerful.
- There was a talk by two young CERN physicists, which was epic. During the keynote, on the big stage, these two guys proved the existence of the Higgs-Boson in real-time using the data from the super collider near Geneva. Google sponsored 25000 CPU cores for this distributed calculation. You were able to see the data being pulled from S3 with at least 200 G/s. An amazing performance showing the power of cloud computing and distributed processing. And it also proves that science can greatly benefit from industry-driven technology as well.
In the end, we want to say that we are keen to watch how Kubernetes will progress and shape the industry as continuing to be one of the largest open-source projects on earth. Thanks to everyone, who made this event possible. It's been such a pleasure!