Cloud architectures come in many forms, but in general they can be broken down into three distinct categories: infrastructure as a service (IaaS), platform as a service (PaaS) and software as a service (SaaS). The difference lies largely in the functionality you hope to derive from the cloud and the degree to which you are willing to move traditional data architectures to either internal or external clouds.
So far, SaaS has proven popular for those who hope to lessen their burden in supplying leading business applications like customer resource management (CRM) or business intelligence (BI), while IaaS has gravitated toward Big Data storage and backup/recovery applications. PaaS has gotten off to a slower start primarily because it centers largely on custom application development and is therefore the most difficult to design and provision.
But that may be changing now that the enterprise has grown more accustomed to the cloud and is more willing to trust it for higher-order, mission-critical functions. According to IDC, public PaaS is heading toward a $14 billion market by 2017, up from an already healthy $8.3 billion last year. That represents an annual growth rate of about 30 percent, but even then PaaS would only cover about 10 percent of the application development and deployment market, which leaves plenty of room for further growth heading into the next decade.
Part of this growth is due to the advent of well-defined PaaS platforms and a heavy push on the part of developers to explain the benefits of the technology for enterprises struggling to maintain a competitive edge in a rapidly evolving data industry. Chief among them is the Cloud Foundry project backed by EMC subsidiary Pivotal, which recently introduced a new release, dubbed Pivotal One, aimed at extending PaaS architectures across multi-cloud environments. The package includes a number of services targeted at key enterprise functions, such as Pivotal HD for Apache Hadoop development, Pivotal AX for advanced analytics and RabbitMQ for open source messaging. As well, Cloud Foundry is rapidly drawing an ecosystem of companion developers like ActiveState, which recently announced the Stackato v3.0 platform built on the Cloud Foundry v2 release with broad support for both OpenStack and CloudStack environments.
The chief rival to Cloud Foundry is probably Red Hat’s OpenShift platform, which recently saw a slew of new functions aimed at improving application scalability and prior-state rollback for both public and private cloud deployments. The company is also building a vendor community of its own through the OpenShift Origin program, which seeks to leverage the already substantial support for traditional enterprise systems like RHEL, while at the same time halving the cost of leading online memory services to mere pennies per hour.
Meanwhile, Google is taking dead aim at the rising demand for mobile application development services with its App Engine platform. The company has launched the Cloud Endpoints feature that provides development tools and other resources for Android and iOS apps. The package essentially provides a complete REST-based back-end infrastructure that handles much of the rote admin, communications and other functions of app development, allowing the developers themselves to concentrate on building advanced feature sets. The company has also created the Mobile Backend Starter, which provides a complete software stack for mobile development, including a data store, authentication tools and notification services.
To date, most cloud deployment has focused on lowering infrastructure and application costs, driven largely by the mantra “Do more with less.” With PaaS, the calculus is quite different. PaaS not only provides a more economical means to foster in-house app development, but also establishes a more robust environment that can lead to greater functionality and even revenue-generating services that otherwise would never have come about.
Rather than simply doing more of the same old thing, PaaS represents the true nature of the cloud: An entirely new way with which to develop, deploy and manage the enterprise data environment.