There's obviously a lot of interest in using Amazon Web Services (AWS) as a cloud computing service, especially when it comes to application development. But the tools that Amazon provides for managing the AWS environment are fairly limited. This not only makes it difficult for many organizations to invoke AWS services, it can also lead to excessive charges because an organization can't easily turn off a virtual machine running on AWS.
The basic issue with AWS is that, as a service, the number of instances of virtual machines running on AWS can quickly escalate. Each of those instances increases the monthly fee for using AWS, so it's easy to see how "cloud sprawl" can lead to a sudden spike in fees.
To address that specific issue, OpDemand has created a service that allows IT organizations to start, stop, clone and destroy cloud applications on AWS with the click of a single button. According to OpDemand CTO Gabriel Monroy, the OpDemand Command and Control platform, scheduled to become generally available this quarter, provides a template approach to managing AWS environments. It's the responsibility of OpDemand, adds Monroy, to stay current with all the changes to the AWS environment, which offloads that task from the OpDemand customers.
Monroy says that it's becoming increasingly clear that once applications are developed on AWS, they tend to stay there once they are rolled out into production. The challenge facing IT organizations is how to manage those applications. Unfortunately, too many IT organizations want to manage those applications with the same manual processes they use to manage on-premise applications versus adopting a more modern template-driven approach to automating the entire process. In the case of OpDemand, that library of templates is built on top of a set of orchestration services that invoke a modeling language that OpDemand created to manage the AWS environment.
Ultimately, cloud computing is all about expanding the number of applications that any IT organization can afford to run. The challenge that comes with that assumption, however, is figuring out how to manage cloud computing environments at scale without having to continually throw more IT personnel at the problem.