You spend a lot of time chatting with organizations about their cloud strategies. What are the trends you're seeing?
James: Customers are asking two big questions right now: First, ‘what’s next in microservices’? Autonomous development teams pushing to platforms with pipelines is so powerful—they want to do more in that style.
Event-based and streaming architectures are bringing real-time capabilities to heretofore brittle and complex batch applications and unifying app and data teams. These used to be monolithic batch/ETL processes few understood and even fewer could alter reliably. New streaming data pipelines can be built with Spring Cloud Stream by the same development teams who already know Spring Boot. We have clients migrating from Oracle Exadata to Kafka + Spring Cloud Stream for data processing. For the first time, their data processing and applications are all on a shared platform, with shared development skillsets.
With more and more of their events available in a streaming platform, customers are embracing our roadmap for Pivotal Function Service. It will allow them to reliably auto-scale many of their event-based services in correlation with event throughput—including down to zero. The event-centric paradigm is a big shift and it's impacting the whole stack.
The second big question customers ask is: ‘should we containerize our whole legacy estate’?
Many organizations want to bring the repeatability and automation of continuous delivery pipelines to everything they do. They want every application team to function in a ‘devops style.’ This isn’t possible without cloud native infrastructure. Cloud native infrastructure is manifesting a huge new set of capabilities. Several major banks want the capability to completely rebuild a data center in the event of a security compromise. They’ve discovered these immutable infrastructure benefits with their greenfield applications, but now they want them for their entire application estate.
Kubernetes has emerged as the runaway choice in cloud native infrastructure. It has the health check, storage, and networking flexibility required to treat containers as general-purpose infrastructure. Kubernetes has the winning container ecosystem. We see important new Kubernetes-first technologies released every day.
We are working with many clients to take a containerized and Kubernetes-first approach to all infrastructure. It's a powerful infrastructure evolution.
We recently announced that Pivotal Application Service is going to be available running on Kubernetes, and introduced new products based on PAS features. What are you hearing from organizations about the news?
James: When we started the Cloud Foundry project at VMware in 2010, the only way to bring it to any cloud infrastructure was to automate down to the Virtual Machine API level. We also had to build our own container-management system to declaratively run, scale, and manage the applications. It was a heavy engineering lift because the idea of declarative and container-based infrastructure was so new.
Nine years later, Kubernetes has become a clarion industry standard for cloud native infrastructure. It's already our recommended infrastructure for broad containerization of IT with PKS. It's natural and exciting for Cloud Foundry to evolve onto this stack since the vision and approach are so similar. The shift frees up our talent to focus on the demands of the application layer and ensures we can efficiently integrate the constant innovation from the Kubernetes ecosystem. A great example of this benefit is our new Envoy on Kubernetes based routing option which brings fine-grain traffic management blue-green deployments.
While these speed and efficiency gains are great for us internally, our customers are also excited. Our platform convergence begins to solve a big gap in the current Kubernetes ecosystem: the application developer experience. It's routine to see Kubecon speakers open with slides reading “don’t expose Kubernetes to developers” and now enterprise developers can get to Kubernetes infrastructure without leaving the comfort of a ‘cf push’ experience.
The best cloud native application and container platform communities are now united. There isn’t a more complete story on the market.
How do you think companies will benefit from the news?
James: Customers can now combine their initiatives in adopting the Pivotal Container Service (PKS) and Pivotal Application Service into one. It’s now natural to start with Kubernetes, via PKS, and layer additional application infrastructure and developer functionality from there. Think back to the two big questions I get—getting deeper on microservices, and broader on containers—and this announcement is a terrific packaging for those needs.
PAS will now become one of the largest enterprise workloads in the Kubernetes ecosystem. With many hundreds of thousands of containers running in production environments, our contributions to Kubernetes are going to dramatically increase and be based on mission-critical enterprise experience. It really is a win for our customers, the cloud native community, and Pivotal.
How should companies think about Pivotal and VMware?
James: Over 70 percent of the Pivotal Platform applications are already running on VMware infrastructure, so in addition to our great engineering collaboration on PKS, VMware is also our top platform infrastructure. I do countless customer meetings in close partnership with both VMware and Dell, and customers love when we show up as the unified cloud-platform company. It's genuinely one of my favorite parts of my job because the combination is so unique in the market.
This milestone allows us to further leverage VMware’s unique engineering experience and capabilities in infrastructure while expanding our focus on developer experience, Spring and microservices scaffolding.
Customers wanted a completely unified and aligned vision and roadmap from VMware and Pivotal. I’m excited about how tightly this move aligns the vision and products of the two companies. You really can start with either Pivotal or VMware now and get the same cloud native solution.
There seem to be two major pools of innovation in infrastructure software where all the great new technology and capabilities are coming from right now: Public Cloud Companies and Open Source Software Companies. Would you agree? And how can Enterprises best position themselves to take advantage of this situation?
James: Public clouds are having to reckon with the power of open communities versus their previously proprietary approaches. The market is highly competitive, and no one cloud is able to set the standard. Developers are flocking to open communities, and the cloud market has responded. This year we've witnessed Amazon refocus its container efforts on Kubernetes and its streaming efforts to Kafka—away from two previously proprietary solutions.
This is important to CIOs. They were never enthralled with being all-in on a single cloud vendor. By taking a Kubernetes-first approach and building applications with the best in open source innovations—like Spring Boot, Kafka, RSocket, and Knative—enterprises can leverage any infrastructure and also hire from the largest possible pool of great developers.
The notion of there being a trade-off between the speed of innovation and open multi-cloud choice has left the conversation. By rapidly incorporating these new open technologies into our platform, we are giving CIOs infrastructure choice without compromise.
We are seeing a big wave of CIOs asking us to partner on this Kubernetes-first operating model for public cloud. They need a partner really focused on enterprise-ready cloud native software to accomplish this. VMware especially has been a huge beneficiary of this trend. Competition between public clouds is creating a renaissance for modern infrastructure software.
So given what you're hearing from customers, and how closely you work with VMware, can you provide a sneak peek at what's to come?
James: If you want to hear more about the future come see my talks at VMworld in August, and at Spring One in October. Function as a service, streaming platforms, reactive event-based microservice protocols and version control driven API gateways all together—the big innovations are happening in the open communities, which we’re a big part of. There is way too much going on to begin to mention in a quick Q&A. This next wave of cloud developer innovation is going to be really fun to watch.
About the Author
More Content by Liz Huang
