![]() |
Infrastructure-as-a-service (IaaS) offers small organizations compute, network, and storage resources on par with the IT capabilities afforded previously to only larger, well-capitalized organizations. Now, PaaS has the potential of taking this IT on-demand concept a step further with applications and services. IaaS was first popularized by Amazon Web Services (AWS). The startup community embraced the model where no upfront capital expenditure was required, and IT burden was significantly reduced because IaaS provided both infrastructure and operations management for a fee. This model gave startups, and larger enterprises alike, a much faster time to market than building out infrastructure themselves. Since being introduced, IaaS has gone mainstream in enterprise organizations. EMC itself is 92% virtualized and provides IaaS to its internal customers. Following a similar course taken by IaaS, PaaS is currently available on the fringe of enterprises from providers like Heroku and CloudBees, as well as more established providers like Google App Engine and Microsoft Azure. Like IaaS before, PaaS could gain significance as the offerings mature and enterprises get savvy to the benefits of this platform model. IT On-Demand What is PaaS? PaaS, like IaaS, offers IT on-demand from a virtual environment that, for the most part, removes operational concerns from the customer, allowing them to focus on core competencies and the business at hand. The fundamental difference between IaaS and PaaS, however, is in the unit of currency. IaaS barters in units of compute, network, and How does PaaS compare to IaaS? To better contrast and compare these two cloud-service models, consider a scenario where developers need to build a new Web application. With IaaS, developers need to:
With PaaS, on the other hand, developers only need to:
Taking IT up a Notch PaaS is not just about developer productivity, but also about operations. With IaaS, cloud services providers set up, manage, and present the virtual servers, networking, and storage, making sure there are enough of these resources available to be consumed in the units associated with each resource (e.g. VMs, bandwidth, and terabytes). With PaaS, cloud services providers still ensure the raw compute, networking, and storage resources are available and properly configured. But, the providers also present the application containers (e.g. application servers and development frameworks) and services to be consumed in the PaaS model. For example, the provider ensures enough application or data base servers to run applications are available to be deployed on-demand. If application server or database capacity gets low, the PaaS provider is responsible for making more of these resources available. What are the important parts of a PaaS? There are four (4) areas to consider when evaluating a PaaS provider:
While still too early to predict, the future of platform-as-a-service offerings looks full of promise—and are potentially as relevant to enterprises as IaaS before it. PaaS resonates with the themes consistent with cloud computing—the reduction of capital expenditure and IT burden, and increased agility—which should continue to make it attractive to many organizations dealing with budget constraints and delivery issues. With PaaS, organizations can focus on higher-value activities like getting core business offerings or services to market faster—even if the market is internal customers. For more information on PaaS provider offerings, read the recent InfoWorld article playfully titled: Which Freaking PaaS Should I Use?
Thank you to Cornelia Davis for co-authoring this post about platform-as-a-service (PaaS). Cornelia is a senior technologist in the Pivotal Initiative at EMC Corporation. Her primary focus is on integration approaches with an emphasis on RESTful SOA. |
Update your feed preferences |
