Thoughts about Cloud Computing and Amazon S3

Cloud computing is one of the rages online nowadays, with tons of applications hopping onto the bandwagon with services like Amazon S3. Wikipedia defines cloud computing as "a style of computing in which IT-related capabilities are provided 'as a service', allowing users to access technology-enabled services from the Internet ('in the cloud') without knowledge of, expertise with, or control over the technology infrastructure that supports them". Basically, cloud computing is storing information in the "cloud" or a centralised location and then accessing it from that centralised location. The best example of cloud computing is Amazon S3 (external link), a service that many websites and online applications use for cheap, centralised, and scalable online storage. Most web applications in a way are examples of cloud computing because the user's data is stored online by the application rather than on the user's computer.

There has been a lot of criticism of cloud computing recently, especially over the reliability of the model in general. For example, if all of your users' data is stored centrally and that central location fails, then all of your users are going to be quite angry. On the other hand, when each user stores their own data on their own computer, the chance of all their data being lost simultaneously is highly improbable. Amazon S3 itself has had numerous outages and occurrences of data loss that have left the Web 2.0 community on its knees. Once Amazon S3 goes down, so do the thousands of web applications based off of it. Sure, the storage is dirt cheap and completely scalable, but are you willing to risk data loss and outages?

Granted, it is impossible for Amazon S3 to provide 100% uptime and they certainly are already more reliable than most web hosts. Personally, I think that cloud computing is the way to go. Amazon S3 is extremely cheap and perfect for scalability. The outages and data loss occurrences are not too numerous and Amazon is constantly improving uptime and performance. It is just a model that is more efficient than conventionally hosting one's own servers. I'm just waiting for a client project that has to utilise mass storage so that I can actually develop on Amazon S3 myself!

If you have any thoughts about cloud computing and Amazon S3, please feel free to post them below! I'm interested to see what others think.


  • How easy is it to keep a back-up of the S3 stuff yourself?

    Posted by Frans (external link) on Tue 14 Oct 2008 at 8:20

  • That would be something I don't know, but I imagine it isn't too hard. I know a lot of companies are now backing up the data in some way after that period of outages.

    Posted by Ethan Poole (external link) on Tue 14 Oct 2008 at 15:10