Using IaaS
Organizations use IaaS for a variety of needs. Some companies want to leverage IaaS to quickly and economically provision additional compute power, whereas others want to avoid costly and complex maintenance of servers. Some companies find that using IaaS for disaster recovery is highly cost-effective and enables the company to mirror its entire site. In hybrid environments, companies will find a myriad of uses for IaaS. A wise user will evaluate each opportunity to use IaaS against the risk level associated with the IaaS provider and the overall economic risks and benefits.
Using IaaS for discrete projects
Here are some examples where IaaS is used for specific projects:
A team at company X is tasked with creating and testing a new application. The team requires a server for testing and development, but management needs the work done quickly and on the cheap. The team can acquire the resources from IaaS provider.
Company Y hosts a very popular food and cooking website. Normally, company Y’s website runs quickly, but before major holidays, the site slows, and last Thanksgiving the site crashed and left cooks wondering what to do with their frozen turkeys. They employ IaaS to deal with any peaks in load.
Both company X and company Y can utilize IaaS to extend their internal IT infrastructure for a short period of time.
These two different companies may each realize similar benefits from using IaaS for discrete projects during a specific period of time. The development and test team will appreciate the high level of control they have for selecting the specific requirements for meeting their needs. The food website company will gain value from the easy and cost-effective way they can extend their IT infrastructure. Here are the key benefits:
Creation of virtual machines with almost limitless storage and compute power
Flexibility to select the developer tools the team wants to use
Flexibility to select the operating system the team requires (Windows, Linux, and so on)
Cost savings from eliminating capital expenditures on large systems that will be underutilized much of the year
Flexibility to dynamically scale the environment to meet their needs
An organization needs to know its potential risk if its IaaS provider experiences downtime while they are using IaaS. A network outage may be a relatively minor inconvenience or even highly problematic for a specific project, depending on the circumstances.
When IaaS becomes the business
Increasingly, some companies aren’t simply relying on IaaS for bursts of extra compute power, but are running their whole business using IaaS. For example, in 2010 Netflix moved to support the majority of its member traffic using IaaS provided by Amazon’s Elastic Compute Cloud (Amazon EC2). Netflix needed to quickly scale its infrastructure to keep pace with the increased demand from a rapidly expanding member base using various new mobile devices. The company used IaaS to run all aspects of its business, ranging from building recommendation systems for members, to storing large amounts of data, to providing backup and redundant environments.
If Netflix had attempted to operate its business using traditional internal data center resources, it’s hard to know whether the company would have been able to scale its infrastructure to support increased consumer demand for its services. Regardless, the company would have needed to accurately predict this increased demand several years prior in order to provide enough lead time to build new data centers.
Whereas Netflix is an example of an established company moving to IaaS, many startups never establish data centers and, instead, open shop from the start using IaaS. Zynga, the online social gaming company, which went public in December 2011, was born on IaaS and continues to stay there.
For Netflix, Zynga, and many other companies, IaaS is not simply a fast, cost-effective way of extending their IT needs; it is mission critical. When a company lives on IaaS, it must take extra care when selecting a vendor.