#cloud forecasting

So, you've heard about the cost savings that cloud based hardware can provide to your organization. But you've been stuck, when it comes to analyzing the available options for cloud infrastructure services. Anyone that's considering a move to an IaaS (infrastructure as a service) model of hardware delivery should have a basic understanding of the associated pricing for such services. Likewise, we're going to fill you in on what you need to know, to successfully forecast the potential costs of an IaaS deployment for your enterprise.

Cloud infrastructure delivers computing capacity for the web applications and services that an organization offers to its users in the cloud. IaaS vendors typically provide a wide range of options for their services, in terms of processing power, memory, operating systems, storage capacity and more. This provides the flexibility that's needed to meet the hardware requirements of different types of businesses. There are a number of solutions that you may want to consider for IaaS deployment, including: Amazon EC2, GoDaddy, GoGrid, Rackspace, Softlayer, Terremark and others.

The majority of IaaS vendors offer some form of pay-as-you-go, unit based pricing that can be ultimately converted to a per hour rate. In addition, some of the most popular cloud infrastructure solutions are priced based on utilization per instance (i.e. cost per hour for a given virtual computing resource such as a server, database, storage disk, etc).

On Demand Instances

Infrastructure services such as Amazon EC2 provide on demand hardware capacity (instances), that are priced by the hour. These on demand instances are delivered on a pay-as-you-go basis that's free of any long term commitment. The ability to pay by the hour for computing capacity greatly reduces the need for developing complex forecasting strategies for cloud deployment. IaaS products like Amazon EC2 provide “elastic” infrastructure services that are designed to meet hardware utilization needs of any scale.

Extra computing capacity for graphics processing, heavy database input/output and network intensive applications are among the options offered. Once you have deployed the desired number of on demand instances for your organization's cloud infrastructure, you can use cloud monitoring services to determine if any adjustments are needed in your cloud deployment, which will help your business get the computing capacity that it needs at the lowest available price.

Spot Instances

Spot instances provide another option for acquiring infrastructure services on a pay-as-you-go basis. Spot instances are bought by placing contingent purchase orders for unused computing capacity in an IaaS provider's public cloud. Orders for spot instances are only fulfilled if the requested resources are available at or below the price set by the buyer at the time their order is entered. Spot instances are appropriate for computing tasks that are not time sensitive or mission critical to an organization and come with the benefit of significantly lower infrastructure costs when used.

Reserved Instances

Reserved instances are another option for acquiring infrastructure services. This type of pricing model combines pay-as-you-go billing with up front “reserve” payments. In the case of Amazon EC2, customers pay a flat fee for each instance they need to reserve, in addition to an hourly charge for the desired level of computing capacity. This pricing structure translates into a lower hourly rate per instance, as compared to on demand instances. Amazon EC2 also provides volume discounts for reserved instances.  Small businesses and organizations who are implementing a cloud based infrastructure for the first time may be better off opting for on demand instances, to save on upfront costs. On demand instances also offer more flexibility for experimentation and change, since reserved instances require a commitment of between 1-3 years with services such as Amazon EC2.

Comparing Pricing Options