While at the Cloud Foundry summit earlier this year I took the opportunity to sit down with James Watters, head of all things cloud at Pivotal, to talk about where the VMware spin-out is heading. Watters has been involved in the Cloud Foundry ecosystem since day one, and has, over that time, been a strong figurehead for the Cloud Foundry ecosystem specifically, and the future of enterprise infrastructure more generally. While Watters and I have had something of a checkered past, with the odd Twitter-flame war for good measure, the parallel rise of the platform he shepherds, the company he works for and his own career has been interesting to watch.
The timing of our conversation was particularly fortuitous given the rumor circulating that Pivotal is looking to IPO next year some time. Pivotal was, as I mentioned, spun out of VMware and gained a healthy investment from, among others, GE when it was formed. It has always been seen (in my eyes at least) as the most valuable jewel (albeit a fledgling one) in the broader DellEMC portfolio, but one which was, at least a little bit, toxic to the broader short to medium term interests of the parent company. DellEMC is, after all, primarily a hardware company, and Pivotal’s approach differs markedly from other vendors trying to foot it in the cloud space (HPE and IBM are good examples). Whereas those two legacy vendors see software as a way of generating greater hardware revenues, Pivotal is “all in” on being a software vendor.
In our conversation, Watters spoke of Pivotal’s strategy with regards Cloud Foundry being about delivering the four different cloud-native abstractions on a single, consistent platform. These four abstractions, as Watters sees it, are:
- Functions (i.e. serverless)
- Applications
- Data
- Containers
I wanted to dive into this four-sided proposition and an announcement at Pivotal’s Spring One Platform event recently closed the loop on Watter’s previously espoused aspiration. At the event, Pivotal announced the latest version of its Cloud Foundry-based offering, PCF. Indeed, with this release, Pivotal reorganized its various product offerings underneath the PCF banner and brought new features and capabilities beyond those of its Cloud Foundry genesis. The rationale behind the move is that, in this new cloud-native era, organizations need to focus beyond infrastructure and look at the totality of their needs. Interestingly, this move very much parallels similar positioning – witness the OpenStack Foundation and its move into the container world.
Anyway – in terms of what the latest version of PCF delivers, here’s a run-down:
- A new serverless computing product called Pivotal Function Service
- The general availability of Google/VMware Kubernetes container product
- The launch of new “Apple App Store-esque” PCF marketplace featuring GitHub, Splunk, New Relic, and hundreds more.
- First-class support for Windows Server 2016 containers. Customers running applications built with Microsoft’s .NET Framework gain access to container features similar to those used in Linux systems. Many PCF features (e.g., CPU autoscaling) now “just work” for .NET applications on Windows.
- Comprehensive security. InfoSec teams can protect customer data better with Pivotal Cloud Foundry. The platform reduces the risk posed by manual practices. Instead, Pivotal customers use an automated “3 R’s approach.” Repair systems quickly, repave environments often, and regularly rotate credentials.
- Network management and security. Integration with VMware NSX-T to help IT administrators manage the container service on their corporate networks. Network administrators can use PCF and NSX-T from VMware to create a common operational model for cloud-native and traditional apps. This helps enterprises create and enforce network access policies across their entire organization.
Essentially what Pivotal is trying to achieve with this new approach is to bring some consistency across all of the various needs that an enterprise has. It’s not just about managing and automating virtualized infrastructure, it’s also about covering container use-cases, serverless approaches and the myriad of third-party tools that are important in the new style of composable enterprise – log management, analytics, version control and monitoring are all required components, so this “app store” approach makes sense and does a good job of positioning Pivotal as the enterprise technology vendor for the new age.
Jumping on serverless as part of the broader offering
In terms of what Pivotal is doing in serverless, while still early in terms of enterprise adoption, serverless is an important emergent trend. Amazon Web Services’ Lambda offering bought the serverless construct to prominence and was quickly followed by Google, IBM, Microsoft and Oracle delivering similar concepts.
However, rather than reinventing the wheel and building a new serverless framework, Pivotal decided to use its existing platform infrastructure to offer a serverless-style lifecycle for more traditional applications. This offering is both the culmination of existing capabilities and an attempt to distance Pivotal from current serverless offerings which have a tendency to force lockin on customers. Similar to IBM’s OpenWhisk and Oracle’s fn, Pivotal is trying to push a function-based paradigm which allows organizations to run the functions anywhere.
Tensions with the parent? Or a smart strategy?
Pivotal’s embracing of both Kubernetes and Serverless are a somewhat stark indication of where it sits in contrast to its largest shareholder, VMware. VMware is, of course, the company that made its money popularizing virtualization but as new, more svelte approaches towards infrastructure have grown to prominence, VMware increasingly feels left behind. Add to that the fact that VMware itself sits underneath the DellEMC mothership and you have a complex juggling act.
All of which makes it even more interesting to consider what the next steps might be for Pivotal. Is Michael Dell looking to spin the company out to pay down some more debt form the Dell/EMC merger? Or does he see Pivotal as a potential future cash cow to replace the traditional revenue which must surely, over time, dissipate?
Whatever the playbook, Pivotal continues to impress with its execution – from a fledgling spinout only a few short years ago, the company has ducked and weaved to perfection and continues to both grab hold of emergent technologies, while still executing more traditional models. Always a fascinating vendor to watch.
