Attending the Pivotal SpringOne conference last month has hit home how important is the alignment around Kubernetes in the cloud-native technology world. This event is a developer conference for the popular Java web framework Spring — Pivotal was keen to quote from the recent JetBrains survey that the two most popular offerings in this category are Spring Boot (56%) and Spring MVC (43%), the next most popular stood at 6%.
This report reflects my experiences at the event wearing Kubernetes-tinted glasses. The reason for this is that to play in the cloud-native world today you need to be part of the open-source Kubernetes ecosystem. The pace of innovation is so rapid that it makes no sense to replicate your own equivalent, whether closed or open source, and try to keep up with the efforts of a much larger community.
Kubernetes has emerged as a de facto standard in cloud-native computing and it has achieved that because it is open source, vendor neutral, and its timing was perfect in solving the need to manage containers. Originated by Google, the open source project is today owned by the Cloud Native Computing Foundation (CNCF), a non-profit, vendor-neutral organization in turn owned by the Linux Foundation. The Kubernetes ecosystem allows many players, from startups to Internet giants, participation in cloud-native computing, fueling its growth and evolution
VMware CEO Pat Gelsinger was invited on the opening keynote stage and talked about the company’s acquisition of Pivotal. VMware gave birth to Pivotal as an external partnership with Dell and EMC (now part of Dell), and bringing Pivotal inside VMware is a strategic move that is all about Kubernetes. VMware recently announced a Kubernetes native vSphere (project Pacific), and project Tanzu — a build, run, and manage offering for cloud-native applications, with again Kubernetes at the center. This shows the pieces of VMware’s strategy falling in place, explaining why the embracing of Pivotal technology even tighter within the VMware stack makes a lot of sense.
All the public cloud players want to facilitate Kubernetes-based cloud-native applications and with VMware playing as the middleware for the cloud, it can benefit in multiple ways: Pivotal gives it the grassroots developers, and its VM infrastructure stack attracts its enterprise customer base to the public cloud players who want to run those enterprise workloads.
The cloud is also a strategic play for Google, and it has played a benign role, supporting the open-source community. So, it came as a surprise that two important open-source projects in the Kubernetes ecosystem, Knative and Istio, expected to join CNCF, will remain managed by Google. While there is nothing wrong about this move by Google, given how much investment is flowing into the Kubernetes world, there will be suspicion that Google will steer these projects towards serving the Google Cloud better than rival clouds. It remains to be seen whether Google can convince the community of the wisdom of its announcement to control these projects or whether the community decides it is better to create a fork.
Moving to the cloud, and re-architecting applications to run optimally on the cloud, is at the heart of digital transformation. Pivotal provides cloud technology but to many of its target large enterprises, the re-architecting element is a huge undertaking. To help enable this is Pivotal Labs, the consulting body that helps enterprises master agile and DevOps and enter the brave new world of containers, microservices and more. Large enterprises carry a huge amount of legacy code (‘heritage’ is a nice term I heard used), and to help transform this, Pivotal Labs has created agile techniques for large system modernization. This is a combination of methodologies and tools, such as the Boris method – a process for mapping architecture components (named after The Who song, “Boris the Spider”), and the SNAP (Snap Not Analysis Paralysis) method, which uses a Pivotal tool called App Analyzer that automatically analyses code and provides complexity analytics, such as degree of monolithic coupling across components. The Labs concept is a huge success, and key competitor IBM Red Hat has emulated it, but is difficult to scale in large multinationals. The path taken by two customers, Dell and a large telco, is to create an internal version of Labs and have Pivotal train the trainers.
Pivotal arranged a customer panel to talk to analysts. My two take-aways were: 1) tools do matter — while culture and people clearly matter, good tools (with first-class automation) make a huge difference in easing digital transformation, and Pivotal Platform (formerly Pivotal Cloud Foundry) plays this role; and 2) finding recruits with cloud-native skills is a persistent challenge. There is a large gap in our educational systems: colleges are not turning out vocationally trained software engineers and so graduate recruits into high tech have no idea what cloud native computing is. Therefore, it’s not just a question of reskilling IT staff whose comfort zone is waterfall; it’s also necessary to train new recruits.
Finally, a major announcement at the event is the partnership with Microsoft on Azure Spring Cloud. In development since February 2019 and going live in early 2020, the arrangement gives developers a seamless and effortless move to the Azure public cloud via Spring; the platform takes care of all infrastructure concerns, providing ‘Spring Cloud as a Service,’ and is a managed service by both companies. Platforms like Spring carry significant enterprise workloads, and as observed above, public cloud providers want large enterprises as customers, so expect to see interest from other public clouds.