3

I am considering moving various VPS servers I mange on behalf of clients to AWS (EC2). However, I would require a fixed monthly cost to estimate what I would charge my clients for the hosting/support up front. Reserved instances would probably not be cost effective as they are often not particularly busy websites - the on demand is what I am after.

My question is, how does one calculate the maximum possible cost of an instance without knowing how busy they may be, or better still is there any way to limit the costs to make sure the client never exceed their pre-paid charge?

Improve this question
3
  • Some clouds do offer pay as you go/on-demand with a spending cap however when you reach the cap they turn your services off until the next payment cycle.
    – Zhaph - Ben Duguid
    Jun 8, 2017 at 20:56
  • I can see Azure does this, as an example. However, it still begs the question - how to use these services to provide a fixed cost hosting to third parties, who will not want their server to switch off, yet require a fixed budget. Many must be doing it, but under what approach?
    – Flash
    Jun 9, 2017 at 1:14
  • As you note, I've got more experience using Azure. In that instance you can use App Service instances and host multiple sites on them, so the cost of the compute instance would be fixed, however what is harder to calculate is the exiting bandwidth costs. As an example you could host 10 sites on a B2 instance for about £80/month and charge each customer £20-£30 a month and know you are likely to cover your bandwidth costs. I've not really got any experience with AWS though
    – Zhaph - Ben Duguid
    Jun 9, 2017 at 6:08

2 Answers 2

Reset to default
1

You should reconsider your requirements. Reserved instances on AWS are better suited for continuous operation of websites. I guess you are trying to minimize the costs of managing traffic spikes without opting for the larger instances all the time. Amazon also offers automatic load balancing. If you want to micro manage switching between instances and stopping unnecessary ones, you will have to use orchestration tools such as Kubernetes. Amazon provides a Simple Monthly Calculator and calculates rough estimates based on all products you wish to use including details like the expected utilisation and bandwidth.

If you don't want to do that yourself, companies like Heroku offer fixed pricing and operate on top of Amazon AWS. They manage everything for you (at a price) and and let you scale your websites on demand.

Improve this answer
1

The major cloud service providers use different pricing models. You could consider at least checking out the big 3: Amazon, Microsoft and Google. Of them all, I think Amazon has the most confusing pricing structure. A lot of that is due to the wide variety of cloud products they offer. You can look at this pricing 101 post to get an idea of AWS storage costs.

Google has been expanding their cloud services. As part of that they offer a wide range of discounts, including sustained usage (longer you use the more the discount) and committed use, where you could set the CPU and RAM per month, with unlimited VM's, for more stable pricing. And they have reserved instance discounts, too. I read this article which has good comparisons and some guidance on which provider is better for what. Plus, in that link there is a cost calculator for AWS and Google.

Improve this answer
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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