Twitter Dies Under DDoS Attack, What’s Next for the Cloud?
Aug/090
If your mission critical apps were part of Twitter’s cloud today, you would’ve been SOL for a few hours while the Tweetsters sorted out a Denial of Service attack that took the system down. Mashable reports that Facebook was also affected. But the news is not that Twitter was down, Twitter actually goes down all the time, just not for hours. At least not recently.
So if your business were running on the Tweet-Cloud (I know it doesn’t exist, I’m just making a reference) and your ecommerce transactions depended on the 99.999 uptime on the brochure, your two hours of dark, in the middle of the US morning, could be quite expensive.
I was talking about Dell and their “cloud computing” strategy the other day a with an enterprise 2.0 advocate and here’s what we came up with.
Let’s say Dell moves their consumer transaction business onto Azure, from Microsoft. And then let’s say there is a problem with something in that cloud, and just for fun (using an actual example that struck Dell over the Christmas holidays) let’s say your webstore went off line during one of your most lucrative times.
I can imagine the executives at Dell getting on the phone and rounding up their IT staff for a WTF meeting. Butts would be in seats, careers would be on the line, and the millions of dollars that would be missed for each offline minute would be accounted for later when the reckoning was taken. In that scenario Dell would call Dell employees and have 100% leverage on their time.
So now, imagine the same Christmas time scenario but this time the Dell executive calls Microsoft and gets the Sr. Manager on duty for Azure. Now it’s not that Mr. Azure is not going to take the call from the Dell VP seriously. Of course all available resources will be put on the effort to get the Dell store back online. And of course an accounting will be made of what went wrong. And of course EVERYTHING will be done to keep it from happening again.
But the leverage that Dell can exert on the Microsoft employee, the Azure Sr. Manger at any level, no matter how grevious the failure, would be significantly less than the heat that would be felt by those Dell employees.
So how can the cloud be held accountable in the same way an employee is accountable? Microsoft could have insurance and pay Dell for any possible interruption of service. Dell could fire Microsoft and take back their store, although once the cloud is deployed it’s not as easy as redirecting web traffic back to Dell’s home servers.
How can the cloud be as responsive as an owned butt in a seat on Christmas morning?
@jmacofearth
permalink: http://bit.ly/tweet-cloud
Cisco Draws a 4 Tier Map of the Clouds and the Future of Cloud Computing
Jul/090
[Crossposted from Uber.la]
The Four Tiers of Cloud Computing according to Cisco.
[Excerpt: from the Register: Cisco cuddles all clouds but one]
Clouds mean a lot of different things to a lot of different people. Warrior explained that Cisco sees cloud computing as having four tiers. The lower tier is an IT foundation, including servers, storage, and networking, and that the whole point of UCS was to be a player for infrastructure. In this area, Cisco plans to compete with IBM and Hewlett-Packard as well as partner with EMC, VMware, Microsoft, and others.
The next tier up is what Warrior referred to as infrastructure as a service, which means selling cloud computing capacity like Amazon does with its EC2 compute utility and S3 and EBS storage utilities. Warrior showed a slide that pegged Amazon, AT&T, BT, HP, IBM, Sun Microsystems Oracle, Savvis, Telstra, and Terremark as the key suppliers so far. And Cisco will not be one of them, even though it must be tempting to build a cloud at cost and sell capacity on it.
Now, taking a step up in the abstraction layer of Cisco’s cloud computing model is something Warrior called platform as a service, and this is really providing cloud infrastructure with software development frameworks that allow companies to deploy applications. This is more like Google App Engine, Windows Azure, and certain parts of Amazon Web Services, and in Cisco’s case, the application framework is WebEx Connect, which is evolving from the online Web meeting platform of the early 2000s into a collaboration framework with APIs for integrating other applications into the Web conferencing, chat, and collaboration tools that can be mashed up as IT organizations see fit.
The top and final tier of the cloudy world that Cisco is helping us all build is software as a service, and here, Cisco absolutely has plans to be a player alongside Microsoft, Salesforce.com, and Google. Up here, WebEx will be the brand. WebEx Mail, a mail and calendaring service based upon the PostPath acquisition from last summer, will be added to the WebEx mix and delivered as a service atop Cisco’s own cloud infrastructure. Dennerline said that WebEx is hosting 220,000 meetings per day and over 4 billion meeting minutes per month and that this was supported from nine data centers around the globe. He added that there are over 450 million knowledge workers on the planet and that the collaboration software and services space would comprise about $34bn in sales and that “we certainly don’t have our fair share yet” of that space. As for how Cisco will get its fair share, it’s the same old mantra: build, buy, and partner.
I think that’s a pretty good demarcation of the space.
@jmacofearth
permalink to uber.la: http://bit.ly/4-clouds
Other cloud posts
- Tag Cloud Computing? – Categories, Keywords, SEO, Tags, Getting to Simplicity
- More About the Azure Cloud – Privatization of the Internet – This Time It’s Personal!
- Cloud Fail – Is The Fail Whale Coming to a Cloud Computing Platform Near You?
- Technorati Fail – The Monster is Out – Another Cloud in Transition
- The Cloud the View and the App (a fable)
- A Tale of Three Clouds – Microsoft’s Cloud Computing Reveals Some Concerns
