Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

Possible Duplicate:
Can you help me with my capacity planning?

I would like to scale a web application to the Cloud and wanted to know if anyone had any experience calculating costs and could tell me how I would go about that. I have NO experience with cloud services at all.

Currently my production environment consists of two web servers and one database server.

If the application continues on a linear growth path, eventually, I want to scale to the cloud to avoid any more long term commitments to extra hardware.

I want to be able to create a similar environment I have now as a baseline. Have this as my fixed cost that I will always have.

I also want to calculate my variable costs that will increase with more users or bandwidth.

I don't have a preferred cloud vendor. Amazon, Rackspace, Terremark or any other is fine as long as understand how to calculate my fixed and variable costs.

share|improve this question

marked as duplicate by EEAA, Michael Hampton, Chris S Sep 4 '12 at 2:20

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

1 Answer 1

up vote 3 down vote accepted

Calculating cloud-costs is simple on the surface, but complex once you get to the spread-sheets.

The simple part is you need to know 5 things:

  • How many servers you need
  • What their duty cycle is (probably 24/7 from the sounds of it)
  • What kind of IO you do
  • How much space you consume
  • What kind of bandwidth you need to the internet (or VPC if you're doing that)

Drop that into a spreadsheet with the cloud-provider's costs and you have your per-month costs.

The complex part comes in identifying those performance metrics, and working out the exact pricing strategy that works best for you. You already know server-count, and that's probably 90% of your total cloud-costs right there; three always-on servers are going to run you much, much more than you're likely to involve in the form of disk-space, disk-IO, and network throughput.

As for pricing, the cloud providers I know do allow different pricing schemes for rarely-on, sometimes-on, and always-on servers. Generally with a variable up-front cost and reduced hourly costs for the sometimes/always variants that come out to a significantly cheaper per-year cost than the straight-up hourly charge.

share|improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.