Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Azure just released pricing a little while ago. And today I got a mail from Amazon advertising their cloud-based CDN.

I feel like cloud-based systems are coming, but not yet ready for mainstream IT systems. I'd like to gain some experience with smaller pilots or short term deployments (like maybe a capacity planning test). What systems do you think are good candidates for moving to the cloud in the near term? (At this point anything with sensitive data seems to be not a good candidate. Am I wrong?) How do you do the cost analysis, both before and after (or do you?)

(as I type this I'm getting a warning "this appears to be subjective.... and may be closed". I just want some input.)


locked by HopelessN00b Jan 22 '15 at 3:18

This question exists because it has historical significance, but it is not considered a good, on-topic question for this site, so please do not use it as evidence that you can ask similar questions here. This question and its answers are frozen and cannot be changed. More info: help center.

closed as primarily opinion-based by HopelessN00b Jan 22 '15 at 3:18

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise.If this question can be reworded to fit the rules in the help center, please edit the question.

we are using cloud for QA and dev but not exclusively. We are testing new features for applications and letting QA test them out. It's a nice playground for now.

That is in line with my thinking. a good way to get to know how it works, low risk, easy to exit. – Cheeso Jul 15 '09 at 3:07

The most sensible thing to do on the cloud that will have a reasonable return would be something that meets the following criteria:

  • Small amount of data needed.
  • Small amount of RAM needed.

Which is blindingly obvious really, Our experiments with the the cloud cost us around $4,500 a month right now and the thing we are gaining is knowledge. It takes around 3 days to transfer the multi TB data set to the cloud and then it takes a fair while to run compute on it. This kind of job is clearly wrong for cloud computing. However we are learning a fair amount around the technology and we can show where its application could be appropriate.

Where does the $4000 go? Is it amazon? I know there's a data transfer fee, but am not sure how it is assessed. 1TB data xfer would probably be pretty high. It does seem obvious: any app that requires significant data upload/download is probably also "wrong" for the cloud. – Cheeso Jul 15 '09 at 11:47
Amazon are the recipients – James Jul 15 '09 at 12:05

We are looking at moving a development and POC/demo version of our system to the Amazon cloud services. As our app runs Oracle and Weblogic the migration should not be too painful.

Developing criteria for ROI is something we are also struggling with. At a basic level I am thinking how to convert the current Data Center costs to per-end-user costs for the system and then I can just check if the per-end user cost of the POC/demo server running on Amazon is less.

You have to include the CAPEX you spent on your DC and equipment as well as the OPEX to make this a meaningful comparison. And if you have to retool the application to run in the cloud then that additional cost must be included. There are a lot of variables...


The compelling business case for the cloud is realized by many IT professionals, but determining the return on investment of a technology such as cloud computing can prove to be difficult. Have you encountered the ITX (IT Cloud Transformation) ROI Calculator? It's a free online service which will enable you to quickly estimate how much you stand to save in the cloud, versus on-premise. The program takes into account software applications, servers, data centers and back office functions and determines which would be cheaper with Web-based software applications and how long it would take to achieve ROI. For more info see this blog post

@we-are-cloud protocol states that you should state your affiliation when self-promoting. That said, it looks like a good resource. – David LeBauer Jan 25 '11 at 21:29

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