5 THINGS THAT AMAZON WEB SERVICES CAN AND CAN’T DO
A sure recipe for disappointment is to expect more from Amazon Web Services (AWS) than it can deliver. While AWS is a rich collection of services that are available in effectively unlimited scale, it’s important to understand that there are a number of things AWS can and cannot do:
-
AWS cannot make your legacy application “cloud-based.” Legacy applications have typically been designed for stable loads with static hardware infrastructure. They will probably work in AWS, but they won’t magically become cloud applications.
-
AWS can support highly scalable applications. Think of AWS as offering infinite capacity. All those applications you had trouble with because they outgrew predicted user load, storage use, or network traffic? No problem anymore with AWS. Amazon provides the resource, you provide the application load.
-
AWS cannot make your application failure-proof. Amazon designed AWS based on the notion that “everything fails all the time.” While AWS is designed to be highly resilient to resource failure, that doesn’t mean your application can’t fail — it just means that you have the ability to make your application more robust, if you leverage AWS application design principles.
-
AWS can make it cost less to run your application. Because Amazon provides AWS on a usage-based cost, if you design your application to follow the “down and off” principle of using only what you need and then skedaddling, you can typically save a lot of money compared to the traditional model of resource cost, where you pay up front for resources.
-
AWS cannot make your application secure for you. In cloud computing environments, security is a shared responsibility. Amazon takes on security responsibility for what it provides — the computing environment — while you take on security responsibility for what you provide — application software components. If you don’t do a good job managing your application’s security, there’s nothing Amazon can do to make it secure.
Thanks to Golden