Just when we thought things were looking up for SMB and enterprise to move into the cloud via Google Apps, Salesforce and distributed file and database hosting, Amazon loses control over their cloud services hosted in Virginia.
It’s now on day two of the outage and really big social media services like Foursquare, About.me and Hootsuite have experienced serious downtime. Countless other sites running using Amazon’s services for ecommerce and product delivery have also been affected.
A big thank you to Arik Hesseldahl on his post documenting a large list of Amazon E2C sites affected by this outage.
Amazon has a status page you can check to see the status of data centers but that’s not making anyone feel better about the situation http://status.aws.amazon.com/
Will enterprise and small business back away From The Cloud?
Unfortunately, events like this outage will cause a resurgence of FUD about storing data in the cloud (Fear, Uncertainty and Doubt). I’ve blogged about downtime from cloud services like Gmail, Google Apps and Twitter before. Whether people are paying for a service or not, when it’s unavailable, they panic.
Very much like 9/11 aftermath, where the NYC tourism and the airline industry were affected, companies who were in the process of exploring, migrating to, budgeting for and planning to move data into “the cloud” may begin second guessing their recommendations to upper management and decision makers.
Small business has quite a lot to gain in terms of saving on infrastructure and technology hires but this may cause many, uncertain of exactly what they stand to gain from the cloud to quietly close the door and stick to expensive, complex on-site information stores. There is the opportunity to bring in a bit of the cloud, a sort of mix and match of internal and outsource infrastructure. The government is trying to move to the cloud as well so we may see setbacks there too.
We’re Sorry, The Web Service You Have Requested Has Been Disconnected Or Is No Longer In Service
What can companies do to reduce the risk of downtime from the cloud?
This will probably go on for about 6 months and then people will forget about this event until the next major outage. Is that the way things should be? Is there reason to fear the cloud?
I think not. The thing to remember is that Amazon provides a service and it’s up to companies who run in the cloud to have contingency plans in place. The term “the cloud” is synonymous with redundancy but being prepared can make all the difference. With VMWare and virtualization of servers reaching an impressive maturity stage, companies may want to consider adding virtualization to their contingency plans.
I attended a VMWare event in Minneapolis a few years ago where a guy shared his case study from working at a major banking institution where they had reduced their downtime from days to hours using VMWare to deal with downtime from natural disasters, terrorist attacks and power grid outages.
It comes down to planning, preparation, being proactive and keeping everyone in the loop once a problem crops up.