Google's Latest Stumble Rains on Cloud Computing Parade
Despite Google's Gmail problems, I am still a fan of cloud computing, I think anything that makes the physical world of computing easier to manage is the future of data processing. I also think the cloud is the future when it comes to how applications are developed and distributed throughout the enterprise. But, thanks to Google, Amazon and others, I am beginning to think computing in the cloud is just a pipe dream. Why would I say that? It's simple, until cloud computing (or SAAS) becomes as reliable as client/server computing, many businesses just won't risk it. Google's Gmail outage only helps to reinforce that belief. While many companies offering cloud computing solutions can attest to the fact that they have reliability down to a science, it is the big names out there that generate the news, and help to solidify people's opinions. In Google's case, one has to wonder which is scarier -- the fact that Gmail went down or what the root cause was. After all, if a well-respected, cutting-edge technology giant like Google can't keep something as basic as e-mail up and running, how can we expect anyone to do it? At least that is the impression that Google is leaving most of us with. For the channel, Google's foibles can be a powerful tool for those still selling client/server solutions -- just mention Google's and Amazon's problems and a shadow of doubt can be placed over the whole hosted applications market -- a great thing if you are not selling SAAS or cloud computing. There are some other sales angles that can work here also. Take, for example, the forthcoming release of Small Business Server 2008, Microsoft's crown jewel of the small-business market. SBS 2008 offers extensive integration options with Microsoft's Live service, which is a family of hosted solutions. Now, if I were selling a customer SBS 2008, I would use the examples of Google and Amazon as a method to keep SBS deployments using traditional client/server technology, instead of handing my customers over to the "Microsoft Live" machine. |
Comments (5)
I do not agree. The client-server model is just as vulnerable and less dependent.
Since the client-server model depends on a company's IT shop to keep the central servers running, patched (for Windows) and connected, that means that they have limited resources to deal with major issues.
I would rather have the full force of someone like Google and many other SaaS company's teams focused on the issues than a limited number of staff.
Yes, there are growing pains, but reasonable companies understand the power of what they get with SaaS versus a siloed approach to managing systems themselves.
Posted by Paul Carney | August 14, 2008 10:42 AM
I have two comments:
1. Sounds like a Microsoft sponsored message. 2. There is no discussion in your article about the frequency of premised based solutions going down or speed of recovery. How many people have experienced their Outlook going down in large and small companies. I think there should be a more balanced approach to your article.
Thanks,
Scott
Posted by scott robertson | August 14, 2008 10:44 AM
Interesting thoughts and not without merit though delivered in extremis. The reader must assume you are making a bold statement because that is a universally accepted method to convey a point. That a large fabric was compromised for a period of time is not, in our estimation, a "sky is falling" reason to abandon an entire method for at least two reasons. First, organization's own data centers and applications fail on a regular basis as evidenced recently with the VMWare "outage". Second, smaller organizations on a shoestring have far greater downtime on internal systems than what we've seen to date from Amazon and Google, so, methinks for SMBs cloud computing clearly remains a "relatively" better value. One final note, SaaS and HaaS (EC2) are absolutely suited to data replication for DR and BC which operate on a store and forward asynchronous basis and offer a great way to ensure data availability.
Posted by Willy WarLock | August 14, 2008 11:31 AM
I'm quite surprised at this posting. I would imagine having this debate three or four years ago, but claiming that SaaS is a pipe dream and recommending to steer away from it in 2008?
It is like recommending to steer away from motor cars and go back to horse drawn carts because accidents happen.
I am sure that that if you check the availability numbers of an average SaaS company you will get much higher results than your average IT department, because that's what SaaS companies do for a living.
Part of the incredible success that SaaS has achieved is precisely due to the fact that the IT was doing such a bad job in providing the service.
Posted by Dani | August 14, 2008 12:10 PM
BALONEY. This is luddite thinking posing as informed journalism.
I suggest the serious IT manager read Sun system's white papers and see where Cloud is going. Utility computing has been relaibel for many years, that's why the majors design and use it.
Posted by Rudolf | August 14, 2008 6:41 PM