Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

PaaS: Between a Rock8s and a Hard Place

DZone 's Guide to

PaaS: Between a Rock8s and a Hard Place

The popularity of Kubernetes has put a lot of pressure on the PaaS market and other vendors in the space.

· Cloud Zone ·
Free Resource

When CloudBees was founded in 2010, some of you might still remember that we were delivering a PaaS in the public cloud — a pretty unique one that was the first to cover the full application lifecycle, from storing code to continuous integration and continuous deployment to our runtime platform. This was well before Docker was even a thing (dotCloud was founded that same year). Then, throughout 2014, we exited the PaaS market and performed the molt that made us who we are today. All this to say that PaaS always had a special place in my heart and I've always kept an eye on the PaaS market and its evolution.

Some of the reasons we decided to exit the PaaS were that the enterprise market (we were pretty focused on enterprise usage) was still slow to adopt public cloud and standardization was lacking: there was no "J2EE" for PaaS, and a lot of companies felt they were locking themselves into the proprietary format of a small company. When Pivotal (with Cloud Foundry) and Red Hat (with OpenShift) started to aggressively invest in that market, we understood that a third "best of breed" option from a vendor with much less deeper pockets wasn't likely to succeed. So we refocused on the core part of our offering: Jenkins, CI, and CD.

Why am I telling you all this? Well, in the last two years, I've been looking with great interest at the emergence and maturation of the Kubernetes ecosystem and it feels like the next natural phase of that evolution is taking place. And this time, the same standardization that pushed CloudBees out of the PaaS market is creating pressure on both of today's de facto PaaS solutions: Cloud Foundry and OpenShift. Kubernetes is the new normal, the new operating system, the "J2EE" (in a good way!) the market has been longing for. This means that organizations that are now ready to aggressively pursue a cloud-native strategy are directly aiming for vanilla Kubernetes, a real standard supported by all vendors, which provides them with a much-reduced vendor-lock in and a great ecosystem of ISVs and SIs.

So where does that put Cloud Foundry and OpenShift? In some ways, with the growing number of Kubernetes-native solutions on the market, PaaS solutions are between a rock8s and a hard place: companies that want flexibility in their deployment can safely leverage K8s directly and companies that want ease-of-use and a low barrier to entry have options with low-code solutions and stateless services. The gap in-between to provide a "Kubernetes++" solution gets tighter by the minute.

The nail in the coffin, if I might say, is that while PaaS vendors need to make money out of their solutions, cloud vendors don't: they provide K8s for free as they are only looking to enable their customers to consume cloud workloads — and that leads to a huge cost differentiator. And with recent initiatives such as Google Anthos, not even the on-premise K8s deployments are safe anymore: with Anthos you get on-premise K8s, fully-managed by Google, for a fraction of the cost of a self-managed on-premise solution.

The next few quarters will be interesting to watch in PaaS-land. Has the depth and breadth of the Kubernetes ecosystem simply killed that market? That's my bet.

Onward,
Sacha

Topics:
cloud ,paas ,kubernetes ,openshift

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}