For CIOs desperate to get to the cloud, VMware just made a play to be their best friend. No, VMware didn’t just release a cloud-in-a-box that magically makes them run like Google. That used to be called “OpenStack,” and it was largely a failure. Instead VMware rearchitected vSphere with Kubernetes inside (dubbed Project Pacific), such that “500,000 enterprise data centers get Kubernetes and a staff of well-trained admins overnight?”, as VMware senior director Jared Rosoff said in an interview.
Or, put another way: Those 70 million vSphere workloads? With Project Pacific they just became Kubernetes workloads. The most interesting part of this move may be how it enables enterprise CIOs to embrace cloud on their terms.
SEE: How to build a successful career as a cloud engineer (free PDF) (TechRepublic)
Of course, the tech is very cool
But first, let’s be clear: Project Pacific is cool technology. As Rosoff has written:
Project Pacific is a re-architecture of vSphere with Kubernetes as its core control plane. To a developer, Project Pacific looks like a Kubernetes cluster where they can use Kubernetes declarative syntax to manage cloud resources like virtual machines, disks and networks. To the IT admin, Project Pacific looks like vSphere – but with the new ability to manage a whole application instead of always dealing with the individual VMs that make it up.
Cool, right? Buried in that blurb are three big takeaways:
vSphere is transformed into a Kubernetes native platform. Developers get a great new user experience, while at the same time vSphere inherits the Kubernetes ecosystem. It looks and feels like Kubernetes to the developer, but like vSphere to the ops person.
App-focused management means enterprises can manage applications as a whole. If an app is 50 components, they can still manage it as a single object.
It all works on the infrastructure (most) enterprises already have. In other words, they don’t have to build up one stack for cloud native apps and another for virtualized apps which, by extension, also means that enterprises don’t need to retrain their whole team.
VMware has, in a way, built an open cloud platform. In some ways, it’s what OpenStack aspired to be, but failed. One reason (among several) that OpenStack failed, and why Project Pacific has so much promise, is that it doesn’t attempt to replace public clouds. Instead, it looks like an ideal on-ramp to them.
The culture of cloud
Push the VMware folks and in their excitement they’ll say things like “What we’ve done is made Kubernetes into the cloud itself” (Rosoff). But I think that’s a bit of a stretch, and not necessary given how impressive Project Pacific is for other reasons, particularly as a training ground for cloud-hungry CIOs.
Consider the fact that Project Pacific gives enterprises the ability to start their journey to becoming cloud native on their terms, as VMware principal engineer (and Kubernetes co-creator) Joe Beda declared in an interview. Cloud, he noted, is really about someone else running something for you with a certain set of qualities (API-driven, self-service, and elastic). It’s possible to have those qualities in spades and yet be completely, culturally incapable of embracing it.
To be cloud native is to have the right organizational know-how, processes, and culture in place to take advantage of those cloud qualities. This is easier said than done, Beda went on: “Companies can move to ‘cloud’ and they’ll spend money and time and won’t see the benefits they expected because the cloud tech didn’t change them into cloud native as a culture.”
SEE: How to build a successful career as a DevOps engineer (free PDF) (TechRepublic)
Enter Project Pacific, which gives enterprises the ability to start a journey of becoming cloud native. They can run Kubernetes right next to their legacy, on-premises Oracle database, their mainframe, whatever. They can start experimenting, VMware vice president Kit Colbert told me, moving to the cloud on their terms:
[With Project Pacific] we can move a customer’s entire fleet of apps forward, giving those apps some cloud and some container benefits, for basically zero – or very low – cost. Of course they don’t get all the benefits of moving to the full cloud/cloud-native/container model. But that’s ok because most apps probably don’t need it. Instead we allow customers to then selectively pick and choose which apps they put the full level of [cloud] effort into, the ones that really differentiate their business.
This “cloudy enough” approach is a good option for CIOs who want the cloud but can’t yet stomach all the refactoring or rewriting of legacy apps to get there. Project Pacific could well give CIOs a middle path that allows them to upgrade all of their apps to a “quasi-cloudy” state, while investing more heavily in the apps that offer the most innovative customer solutions.