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.

I have a pure web service application layer using .NET. I was originally planning to use Amazon EC2, but have recently been evaluating Rackspace Cloud Sites as a potential solution.

The app uses Amazon SimpleDB as the database layer. As a test, I have the app running successfully in Rackspace Cloud Sites. Performance seems to be equal to (if not better than) a standard EC2 instance, even with the added latency of the SimpleDB requests travelling to the Rackspace network. However, testing at this stage is at a very low scale.

My question is this: has anyone had real-world experience running a high scale application on Rackspace Cloud Sites? Moreover, once you pass the "included" 10,000 compute cycles per month, does the overall cost seem to be lower than rolling lots of EC2 or Azure instances? My assumption would be that with completely smooth scaling (i.e. only adding compute resources as needed rather than spinning up full instances), the cost could be lower on average. However, their stated goal of calibrating 10,000 CCs as a single 1.2 Ghz CPU seems on average to be much more expensive than EC2.

I like the idea of no-touch scaling, but is it too good to be true?

share|improve this question
add comment

closed as not constructive by Michael Hampton, Jenny D, Scott Pack, pauska, Ward May 22 '13 at 6:58

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.If this question can be reworded to fit the rules in the help center, please edit the question.

3 Answers

"millions of requests per month" is not high scale.

10 million requests per month at a constant rate works out to 3.8 requests per second. Given roughly average distributions between day and night, means your mid day peak is going to be in the teens of requests per second. That is not even remotely high scale. A single server could handle that, even if its pretty bad code.

share|improve this answer
    
In reality the scale could easily be 250 million/month and up into the billions. I'm concerned about what might happen if things really start to get going, and if Rackspace can actually handle it. I'll remove that bit from my question as it's not particularly relevant. –  Allara Dec 21 '10 at 22:55
add comment

.NET autoscaling is the strong point of Windows Azure. Have you looked into it?

Servervault was running on Azure for a while ;-)

share|improve this answer
    
The Azure web site(s) are just a mess. I'm having a really hard time determining what exactly it does and how much it costs. It didn't look like they did any autoscaling. Let me look again -- thanks. –  Allara Dec 21 '10 at 22:23
    
Looking at it again, I'm not seeing no-touch autoscaling for .NET web apps. They appear to offer only individual instances (just like EC2 or Rackspace Cloud Servers). If I need to manually spin up and down instances, or write scripts to do it, I'll stick with EC2/Rackspace. –  Allara Dec 21 '10 at 22:43
add comment

At this writing, the answer is yes: RS Cloud instances consistently provide more CPU cycles per instance compared to comparably priced EC2 instances. Not just more cycles: A lot more cycles. I am not surprised that you see significantly better perf on RS cloud compared to EC2. So do I. (I see favorable perf on RS cloud compared to dedicated vmware instances!) RE Scale: I cannot speak to "cloud sites" as I have no experience there. I can say that cloud servers can handle a large workload, and you just add instances to scale up.

share|improve this answer
add comment

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