tags:

views:

37

answers:

1

Amazon offers two instance types on EC2: 1) On-Demand and 2) Reserved. After reading the docs on these, I don't really understand the difference from an end-user perspective. More specifically, I'd like to know the answer to this question: is one or the other better for web applications?

Based on their names and descriptions, it seems as though on-demand instances may get wiped away from the server altogether if they're not in use which means that they need to be restarted when a request finally does come in. That seems like a pretty bad thing for a website. Am I just misinterpreting the docs?

Thanks!

+2  A: 

A reserved instance you have to pay some cash up front for either a 1 or 3 year period. This allows you to run the server at a reduced price (about half price per hour). Other than that, there's no difference regarding how it works, if it'll get wiped or anything. So if you expect to be running a server 24/7 for a year or longer, a reserved instance would be a cheaper solution for you.

Xorandor
Actually it is possible to have your instance wiped if it's turned of for some time. It depends what you choose for "root device type". EBS will be sustained where as the other solution (instances I think they're called) are wiped. They're designed for this ramp up/ramp down concept.
Xorandor
Amazon has this to say about their reserved instanced: "Many applications require servers to be available at all times. Reserved Instances could provide your business substantial savings if you currently own your own hardware or exclusively use On-Demand instances for your steady state applications."By this statement it seems there's something more. Are you sure they're the same?
Josh
They are still the same. It's only a matter of pricing/spinning up more machines. When they are reserved you get a lower hourly price and if you have applications running 24/7 or "steady state applications", it's more profitable choosing a reserved over an on-demand
Xorandor