Its cornerstone is Cloud Foundry, an open source, platform-as-a-service available in the cloud on top of Amazon Web Services and now in a downloadable, supported version in Pivotal One. In addition to a multilanguage development environment (Node.js, Ruby, Java, .Net), the platform comes with four services that might be considered the basics of future large-scale applications: Hadoop big data handling, an analytics service, messaging, and a relational database service.
The Pivotal One Platform "represents an interesting intersection of new, more agile development techniques ... We've added the ability to deploy wherever you want. It allows the enterprise to have this same, elastic infrastructure" as public cloud suppliers, such as Amazon, Google, and Yahoo, said Dave Menninger, head of business development and strategy at Pivotal, in an interview.
Although Cloud Foundry, as open source code, has been available for free download, the release of Pivotal One marks the first time a supported, on-premises version has been made available, giving companies the option of using Cloud Foundry for in-house development. That version is now referred to by Menninger as Pivotal CF. The Baidu search engine, the Chinese-originated equivalent of Google, handles two billion requests a day and is built on Cloud Foundry, Menninger pointed out.
[ Want to learn more about Pivotal One's origins? See Pivotal CEO Maritz Teases Linux For The Cloud. ]
That means applications developed with the platform, frequently Java-based, can be deployed as virtual machine workloads to Amazon Web Services; to a private cloud based on VMware's vSphere and vCloud Suite; to a public cloud based on vCloud, such as Bluelock; or to an OpenStack cloud, including Rackspace and HP.
In addition, applications developed on Pivotal One can have data sources bound to them through the platform's automated plumbing.
The four services added on top of Cloud Foundry are: the MySQL database, the RabbitMQ messaging service for exchanging messages between applications, Apache Hadoop, and Pivotal AX, which provides automatic analytic instrumentation.
In addition, third parties are offering buildpacks that allow other software to work with the Pivotal stack. IBM has produced the Liberty WebSphere Buildpack to give IBM customers the option of deploying applications to a WebSphere application server instead of Cloud Foundry's reliance on Tomcat. Such an application would have use of the Hadoop, Pivotal AX, and RabbitMQ services.
"This is the beginning of a market. We're changing the way organizations deploy applications," Menninger claimed. He's referring to an age where it's common to rapidly build large-scale applications using agile development methods. Once developed, they can be deployed from the platform on which they were developed to a variety of cloud services, or they can be easily modified to serve a variety of mobile devices.
Menninger said SAP's CTO Vishal Sikka, in his keynote at SAP's TechEd Conference earlier this year, said SAP is watching the Pivotal One Platform evolve and will consider how to make use of its application automation features.
The crucible of cloud, big data and distributed computing is hell on systems. Will application performance management cool down complexity — or just add fuel to the fire? Also in the new, all-digital APM Under Fire special issue of InformationWeek:Cloud industry heavyweights discuss the pros and cons of OpenStack support for Amazon APIs. (Free registration required.)