RightScale Blog

Cloud Management Blog
RightScale Cloud Analytics Free Trial
Cloud Management Blog

How to Manage Cloud Costs and Optimize Resource Usage

Last year RightScale announced the acquisition of PlanForCloud, a cloud cost forecasting tool that enables you to model your cloud infrastructure, run it through a simulation, and get a three-year cost report showing what your deployment will cost. We know how hard it can be to figure out cloud costs and optimize your usage, so we made it easier with our free cloud cost forecasting tool.

We’ve since added a number of new features, which I’ll highlight here. Plus I’ve got some tips on how you can use this powerful tool to plan and budget for your cloud costs.

1. Filter Your Cloud Resource Options by CPU Count and RAM

When it comes to choosing cloud resources, you have many different options from various cloud providers. Resources fall into three broad categories: servers, storage, and databases. To make it easy for you to compare and choose the right resources for your needs, you can filter using the CPU count and RAM that your application requires and compare different resources, as shown in this example:

PlanForCloud server cost comparison When comparing the costs of various cloud infrastructure providers, you have the option to filer by CPU count and RAM.

2. Get the Full Picture of Your Requirements, Not Just Your Servers

We have talked to many users who began using the cloud with the assumption that hourly server costs were all they needed to consider. But when they got their bills, they were shocked to discover that there were many other costs that they had not taken into account. With PlanForCloud, you can avoid being unpleasantly surprised by modeling all the major components of your deployments including:

  • Server running hours
  • Storage costs
  • Snapshot costs
  • Read and write request costs
  • Archiving costs
  • Database running costs
  • Database transaction costs (IO)
  • Data transfer costs (including within a deployment and outside of a deployment)
     
PlanForCloud three-yearcloud  cost report The PlanForCloud three-year cost report breaks down your cloud deployment costs per month.

3. Compare Purchase Options and Clouds

PlanForCloud includes prices for AWS, Rackspace, Google Compute Engine, Windows Azure, and SoftLayer. Many of our users are attempting to determine if they should buy AWS reserved instances rather than on-demand instances and, if so, whether those reserved instances need to accommodate light, medium or heavy utilization. When it comes to larger reserved instance purchases, you need to look at the upfront purchase costs versus lower monthly costs. To preserve cash flow, many larger consumers choose to stagger their reserved instance purchases to better cope with the upfront purchase costs.

PlanForCloud cost breakdown for AWS This PlanForCloud cost report shows pricing options for AWS, including the cost of on-demand vs. reserved instances.

4. Forecast How Much Your Growth Will Cost

One of the main benefits of the cloud is the ability to increase your cloud usage and grow. However, this growth comes at a price. With PlanForCloud, you can use growth patterns to model your growth to see how much it will cost. There are two types of growth you can model:

  • Temporary Growth: A temporary growth pattern is one in which you see a peak in usage, but after the peak is over, your usage comes back down to normal. Many of our users prefer temporary patterns to forecast their costs for occasions when they issue press releases or launch marketing campaigns, to prepare for seasonality-influenced traffic spikes, or to design one-time experimental loads hitting their applications.
  • Permanent Growth: A permanent growth pattern occurs when your usage of the cloud is constantly increasing. For example, as your user count grows, your app servers increase, and your storage and database requirements also grow.

You can always use a mix of both temporary and permanent patterns to model your growth requirements. In the example below, we are doing what-if analysis for MySQL/Amazon RDS based on hypothetical database requirements:

Using PlanForCloud to model growth patterns for cloud resource usage PlanForCloud growth modeling provides comparative costs for such resources as application servers, storage, and databases.

 

5. Import a Snapshot of Your Cloud Deployments Into PlanForCloud

Many of our users already have proof-of-concept projects or applications running in the cloud. We have added the ability to import your existing cloud resources into PlanForCloud to forecast your costs and compare your options. This feature is currently enabled for AWS only. However, we will soon offer the capability for you to import cloud resources from other cloud providers as well as from RightScale.
Import from AWS and forecast future costs
To use these new features to help you better understand and forecast your cloud costs, simply sign up to use PlanForCloud (it's free) or log in as a guest.

And to keep up to date with all the latest on PlanForCloud news and features, follow us on Twitter @PlanForCloud. We would also love to hear your feedback and feature suggestions via our PlanForCloud feedback portal.

Comments

What we need is a forecast tool that investigates current on premise applications and calculates equivalent cost in the cloud. No one can predict how many transactions and bytes received/send an application will generate.
Posted by L. Blom (not verified)   Ι   January 18, 2013   Ι   06:43 AM
Great post Hassan. Flexible predictions based on the current cloud usage sounds fantastic. Well done, I will be tracking your progress.

Post a comment