Seattle based Amazon, the online retailer who is more talked about these days for its Web Services aka Amazon EC2 cloud computing, had a massive outage last week due to hardware failures triggered by power loss. As the whole world seems to be operating from the cloud these days, the Amazon malfunction again elicited debates about the wisdom of jumping to the skies to cut costs. In this day of 24x7x365 business and super fast service, how much are you willing to rely on outsourced platforms to lead your disaster recovery measures?
Cloud computing is also served up by vendors as a more green option. Highly specialized energy saving environments that are carefully cultivated by huge companies like Amazon and Microsoft are likely to waste considerably less energy than the clients who seek their cloud services they say. But the big elephant in the room when it comes to cloud computing continues to be security concerns as you let go of sensitive client data over an open network. Remote computing increases the risk of breaches. This was recently brought to the forefront with Google’s well documented security breach with Google docs when user generated documents saved on Google Docs Cloud Computing Service was revealed to users of the service who lacked permission to view the files. Certainly Google has been exceptional in embracing green computing and proclaims to run the most energy efficient data centers in the world. But are we willing to ease up on the security of confidential data to get our carbon limits under control?
Even as such questions are being raised, cloud continues to surge ahead. Market research firm IDC recently came out with a study predicting that the marketplace would radically transform in 2010 driven by among other things a growing impact from the cloud services model. IBM launched the Cloudburst cloud appliance for 'private cloud' deployment this year, Salesforce went for its piece of the cloud with Force.com. Amazon’s EC2 and Microsoft’s Azure are constantly updating themselves to keep up with all the cloud moves. Organizations such as the Cloud Security Alliance, comprising of industry leaders, global associations and security experts, have published guidance to come up with secure cloud computing practices and has released guidelines that cover 15 security domains, ranging from computing architecture to virtualization for organizations. Just like any other emerging technology, there are certain matters that require meticulous ironing out.
I’m optimistic that we will be able to make cloud work as we cannot ignore the business benefits that it brings us (lots of cash save) while keeping our individual and collective carbon limits under check.
Amazon.com, Inc. (NASDAQ:AMZN), a Fortune 500 company based in Seattle, opened on the World Wide Web in July 1995 and today offers Earth's Biggest... [more]
We use Amazon EC2 to host http://www.billingboss.com and we did not notice an outage last week. I want to point out that outage is also a possibility with traditional hosting company. Rackspace had a major outage this June that took out a lot of websites. http://www.techcrunch.com/2009/06/29/yes-rackspace-is-down-and-so-are-many-of-your-favorite-sites/
Even if you do hosting in house, the possibility of outage is quite high too unless you have a deep pocket and a well established support team.
With Amazon EC2 hosting, it is a paradigm shift in how you design your hosting infrastructure. The correct way to use cloud hosting (especially with Amazon EC2 hosting where every server instance you run is virtualized) is to assume that everything will fail at some time and plan for it.
Unless all data centers of Amazon go down together (considering there are three of them in the US, two in the east coast and one in the west coast), there is always a way for you to self correct. Amazon EC2 gives you everything you need to self serve. If a server in a data center goes down, you launch a new one in another data center to take its place. This is of course only possible if you have configured your servers to be relaunch-able. (meaning that you can configure a virgin server using scripts with no human interaction). I will challenge any traditional hosting companies that can provide you with new hardware to replace a dead server within minutes. This is the power of Amazon EC2 gives you. Instead of waiting for a support person in a hosting company who is probably overworked and always pressed for time to help you in the event of a disaster, you have the tools to control your own destiny.
Richard, I think your correct I looked for something on a Amazon downtime and could not find anything. And the Rackspace downtime had nothing to due with their cloud offering, power problem took out the dedicated machines as well.
