Hacker News Reader

Comment by onlyrealcuzzo | Hacker News Reader

Almost 80% of IBM's profits come from its cloud division, which is only 27% of it's business.

It seems like the idea behind the Red Hat acquisition would be to increase the cloud division revenue and keep roughly the same profit margin, which in theory could be great for their bottom line, especially if Red Hat is growing quickly (is it?).

The arm chair CEO in me doesn't see how this adds up, though. Red Hat seems pretty pricey. I just don't see how Red Hat competes with AWS, Azure, and GCP moving forward, and I don't really see how it plays nicely / fits into those clouds either. I also don't see how it would make IBM's cloud more attractive. I can see how it could make IBM's cloud more profitable, but it seems like a stretch, given what they paid for Red Hat.

View on HN

Red Hat employee here, working in the cloud division. I can't speak for IBM, I'm not too familiar with it. I don't know what the future will hold.

So far, Red Hat plays in a field slightly different than AWS/GCP/Azure. Our customers mainly want something _on premise_, something on their machines, their datacenter, behind their firewall. Most of our customers don't want to, or can't, hold their data on machines they don't physically own.

I just don't see how Red Hat competes with AWS, Azure, and GCP

It doesn't. Their cloud tech, OpenShift, runs on top of the others. Kubernetes can sit on top of that. It makes private clouds portable (kind of).

Slight nitpick. OpenShift, since its latest major version (3?) is Kubernetes. It's mostly a repackaging/distribution of k8s. OpenShift can run on top of AWS/Azure/GCP, but also, most importantly, OpenStack.

Another nitpick, openshift is not a repackaging but a fork of Kubernetes pure and simple, if you look at the origin you'll see those merge patch they do on every Kubernetes releases

I stand corrected, thx.

So kubernetes turn the clouds into a commodity. And this is Red hat play. The issue here is that google (kubernetes creator) is afraid to compete with itself (aka GCP), and hence needs another player to complete the picture.

Based on market share, Google has undoubtedly run the numbers and realized that increasing cloud portability hurts Amazon and Microsoft more than it hurts Google.

From that perspective, community adoption of k8 benefits Google.

This is correct, however, kubernetes is a platform of platforms. I.e. it will hurt all of them at the end.

Kubernetes is realy a free, uncontrollable , widely installed (on prem and cloud) infrastructure, that compete directly with the clouds. Think of the iphone app store, without any apps (yet), and not controlled by apple.

The reason that companies pays for cloud services (which are 20x the real cost of hardware) is to off load the human cost in managing their infrastructure, and to get hold of cloud native infrastructure (I.e. infra as code, fully scalable).

However, with kubernetes you can create your own automatic robots (aka operator) which would replace the costly admins by taking care of application management. And at the same time you architect to a portable API layer.

We use kubernetes on my team, so I'm familiar with the FAQ. ;)

Reply on news.ycombinator.com