LinkedIn Twitter Facebook
Cloud Computing Evangelist, Blogger and Lecturer at IAmOnDemand.com. Ofir has extensive experience helping ISV companies with cloud adoption and management. Today, Ofir is a Senior Vice President and Chief Evangelist at Newvem. Prior to this Ofir led ClickSoftware's On-Demand initiative and established the company cloud offering. He also held several positions at Zarathustra - SaaS development, including the company CEO.
The thoughts are his own and don’t represent anyone or anything he is currently associated with.

4 responses to “Amazon Outage: Is it a Story of a Conspiracy? – Chapter 2”

  1. 3 takeaways from Amazon’s latest cloud outage | SmartPlanet

    [...] week, taking down popular websites and throwing the cloud community in a bit of tizzy. Since then, experts have weighed in on what people should learn from these outages, and I’ve aggregated a few of the lessons here. [...]

  2. Cloud Servers

    Setting up a redundant or fail over site (Load balancing) could best serve sites that are mission critical. I agree with you, if it happened once, it could very well happen again. We just need to be best prepared for it.
    Good pots though and followup from last year!
    Thank You!

  3. Ofir Nachmani (@IAmOnDemand)

    “It is baffling how Amazon have these issues for such sustained periods, it seems like a strange occurrence that 2 substations should go offline at the same time, but it is possible. When this does happen, a transfer switch in a Tier 3 data center will draw the load from the UPS system (like a room full of car-batteries in the big facilities). A UPS should hold the load for at least 30 minutes whilst the generators start up. Generators should take between 5-15 minutes to warm up and why the UPS plays a simple role in preventing power loss.

    This isn’t a ‘perfect world’ scenario, I’ve visited around 20 data centers on the East coast and Europe and this is simply standard practice, there are some variants but the pricinciple is the same, if the ‘grid’ fails, a UPS carries the DC until the generators are ready.
    It seems like an awfully unlikely set of events and failures that indicate they don’t test these systems until they fail and because of the coverage these outages receive, it’s giving the rest of the industry a bad name.”

    http://www.iamondemand.com/post/25561096171/amazon-outage-is-it-a-story-of-a-conspiracy-chapter#disqus_thread

  4. Charels Kelly

    So, will companies or individuals ever rely completely on the cloud? Is it responsible or will it ever be a responsible thing to not have everything you use and create operating and backed up only on the cloud?
    There will be failures and then there will be a long stretch where there are none and it will be declared that the cloud is as safe as it gets. So, will it be responsible and “best practices” to rely completely on the cloud? I propose that to ever rely completely on the cloud is not responsible and will never be responsible.

    Of course at some tipping point, it will be difficult if not impossible NOT to rely on the cloud. Just as we now rely on the internet. What would happen if the internet (the entire internet, worldwide) went down for three or even five business days? What would happen? It cannot happen though… right?