Storm Clouds on the Horizon? - Our Take from Kubecon

Our love for Kubecon is well documented. When there, we are among friends, partners, customers and prospects - this year was no different, and we had the best conversations we have had at any show this year. The Kubernetes community and MinIO grew up together and thrived together and we remain bullish on what we can achieve together.

There are, however, storm clouds on the horizon. Some of them are a function of how the Linux Foundation/CNCF runs things, others are the residual effects of the massive amounts of capital plowed into startups over the past few years. Nonetheless, there is a trend here, and we have documented it from show to show. This is the Detroit 2022 take:

  1. Detroit was not a great place for the show. Despite superb weather, no one wanted to be there. The Detroit effect may have been compounded by the mask mandate. This was an issue with almost everyone we interacted with. It is not clear that people didn’t come because of that, but they were not happy about it. Either way, attendance was down. They even removed seats from the keynote on day two to cover for the attendance issues. I suspect a lot of the “attendees” were from the vendor community which leads me to my next point.
  2. There were nearly 280 booths on the show floor. They were really spread out (unless you were a Silver Sponsor and were doubled up), but since there wasn’t much traffic on the show floor that allowed us the opportunity to take several laps and talk to a lot of folks.

    We may be looking at a show that is past its peak. The problem was that peak was back in 2019 in San Diego. COVID may have taken the best years.

    Either way, there are way too many companies chasing way too few viable opportunities. The industry doesn’t need 15 API management companies, 15 security plays, 30 observability plays, 15 cost control companies. If you were there you know what I am talking about. Maybe 30% are viable. Of those, many are features that will be absorbed by acquisition into larger players. That does not equate to a bad outcome - but 70% of these companies are in trouble. Everyone sees it. They may have raised last year, but they are now paddling upstream while simultaneously taking on water. My prediction is that Kubecon has 220 or less booths next year. If I am wrong, I suspect it will be because they downgraded.

    One last note on the show floor. It felt very sterile - and empty. I think this is part design, but also part of the way you buy. It was just a long row of booths with different graphics. I think that there needs to be some room for personality to come through - without having to step up to platinum.  
  3. There was very little buzz. Yes, Kubernetes has matured. This is a good thing and what happens when something becomes “enterprise” but it really felt dead. This was apparent in the halls, the keynotes, the side sessions. Something is off. Kubecon used to be a celebration of technological achievement - of something meaningful with a shared purpose. We can’t put our finger on it, it could be a number of things but there is no energy. We were involved in a few pre-event things - there was excitement heading into the event but it never made it into the building.

  4. Part of it is that there aren’t that many “interesting” things going on. Service mesh was cool and some good work was done there - but that was a few years back. The new incubations are smaller and smaller from an opportunity perspective and thus less exciting. Security is imperative and there is good work getting done there but it isn’t really that interesting.

It bears mentioning again, it is difficult to put one’s finger on the source, I suspect that it is a combination of the above and probably some others that we missed, but it is real, I can assure you that. One conversation suggested that OpenStack was briefly big - but we don’t buy the comparison.  Kubernetes is akin to the Linux wave, although others suggested that Kubernetes is but a component of the overall cloud native play. Again, we would opt for pillar - we see it as one of the most important pieces of the cloud operating model.

One clear theme was multi-cloud. It is in everyone’s messaging and everyone’s slides. It makes sense. Every enterprise is a multi-cloud enterprise, if not today, very shortly. Kubernetes is fundamental here and it will ensure the medium term relevance for the technology.

On a slightly different tangent, one area that needs more attention is the data space. While there are some vendors at the show, storage and analytics are underrepresented. It may be that the Snowflakes, Dremios, Starbursts of the world see themselves more as consumers rather than active participants, but we still feel like their presence would add some real world applicability to the event. We know why there aren’t any storage providers - most of them are appliance vendors and that is an anti-pattern in the Kubernetes world. Still, data storage isn’t getting its due - the entire ecosystem is built on data, it has to live somewhere and that somewhere needs to be a first class, Kubernetes citizen. That, however, is a different, longer post.

Net-net - we loved seeing everyone this week, even if we couldn’t see each other’s smiles. We have work to do to inject a little more energy into the community and we will take that on. Look for us to go a little deeper on Operator in some coming posts to support that. Until Amsterdam - stay calm and Kube on!