AKF Partners

Abbott, Keeven & Fisher PartnersPartners In Hyper Growth

Rules for Surviving an Amazon Outage

Because of recent issues with Amazon’s services there is a lot of interest in why some companies are able to keep their site up despite their IaaS or PaaS providers experiencing issues. Here is an InformIT article we wrote, outlining a few rules for surviving an Amazon or other cloud provider outage.

Comments RSS TrackBack 4 comments

  • Ori Lahav

    in June 12th, 2011 @ 11:24

    Hi Mike and Marty
    I read the article and of course can agree to most of it.
    I would somehow challenge your last point of “Be Wary of Scaling Through Third Parties”. all is right and doing stuff through open source infrastructures is the right way to build highly scalable internet applications but… what is AWS? isn’t it a third party? isn’t it bound to fail?
    as much as I appreciate AWS product availability zones are not the answer and I saw AWS completely failing before and they will also in the future. counting on availability zones at this time is like the US airport checkpoints asking you to take your shoes off. it’s closing yesterday’s fault and not the one that will happen tomorrow.
    If you really want to take your destiny in your hands – don’t do it on any cloud!

    one other aspect that you did not talk about is cost. Serving cost efficiency is a great parameter in business scalability. $$$ that you put in serving is going down the drain, it has no incremental value to the business. this is why you have to be very cost efficient and use your dollars on developers & BD people that have great incremental value.
    AWS is insanely expensive and instead of making things scalable it drains the business from $$$ and doesn’t let you grow.

    I’ll be happy to share more thoughts.


  • fish

    in June 12th, 2011 @ 21:12

    Hi Ori,

    Thanks for the great comments. I agree with you that AWS is just another vendor that you should try to avoid relying on completely if possible. We talked in a recent post about the idea of redundant clouds, http://akfpartners.com/techblog/2011/05/09/federated-cloud/.

    Ideally everything should be redundant with two different datacenters/clouds/collocations from two different vendors. I believe this is how you’ve setup your environment.

    We’d love to hear more about your architecture and cost, I know you’ve done a ton of analysis and work to make your environment cost effective and stable.


  • Karla Heap

    in June 13th, 2011 @ 00:04

    This is certainly a terrific article. Thank you so much for bothering to detail all this out for all of us. It really is a great guide!