2

I want an unbiased approach towards cloud term that is being hyped today. Is this really a new term, new dimension or is it just a buzz phrase to re-market existing technologies. Personally I think that the term is just being used to create a new in the market and nothing else. The technologies have been existed since so long. Isn't the cloud the virtualization of exiting resources? I would really appreciate if anyone has a good answer and can change my opinion :)

3

7 Answers 7

4

There is no way to tell whether Cloud Computing is a revolution or just a fad right now. It could be either. It is absolutely a buzzword right now, and as such suits often misuse it totally. But that doesn't preclude it from being important.

If you've read Clayton Christensen's "The Innovators Dilemma", you know how new technologies may be inferior to existing solutions at first, and then reach a tipping point where they become 'good enough' and overthrow the old.

Cloud computing might be on the way to replace traditional on-premises, self-hosted software in a lot of places. Not completely replace, mind you, but take a large or possibly dominating market share. As all past revolutions, if this change happens then it will take decades to complete.

For a quick overview of the arguments in favor of Cloud Computing, I would encourage you to read "Above the Clouds: A Berkeley View of Cloud Computing". It may be a little fanboy'ish, but it pretty much covers the economies of scale for large datacenters and business benefits in Cloud Computing.

In short, some of the common arguments given for Cloud Computing are (and I'm not going to debate them individually, that would take too long):

  • Large economies of scale in operating large datacenters lead to lower cost per unit for storage, CPU hours etc on Cloud Computing.

  • Cloud datacenters have predictable growth, seasonal demand differences between cloud customers, and different peaks in demand amongst different customers. This should lead cloud providers to operate their DCs at higher average utilization than on-premise DCs. This is pretty big, as most on-premise DCs run at only ~12% average utilization. This again should make Cloud Computing more cost-effective.

  • For Cloud customers scaling out can be near-immediate, as opposed to, say, a 1-2 months lead time for purchasing new colo space and new servers. This is in itself a pretty big business benefit.

  • For cloud customers, computing moves from being CAPEX-heavy (buying servers etc) to be more OPEX. Some large businesses like this a lot better, as it 'improves' their balance sheet. Small startups like this a lot, as it reduces and defers liquid cash needs, and startups are always short on cash..

  • For cloud customers, Cloud Computing does (at least in theory) offer the advantage of closely matching resources to actual proven load, instead of anticipated load. Potential cost savings again.

The proponents of Cloud Computing believe that IT may see a change very similar to what happened to electricity. At first electricity was created using on-premises generators. Then the distribution technology (high voltage long-haul wiring) improved, and ultimately almost everyone switched to centralized production of electricity.

So the proponents believe that as Internet capacity and speed improve, centralized 'production' of CPU time and storage will be more desirable than on-premises production. Time will tell if they're right. But some smart people at Amazon, Google and Microsoft and other companies have crunched the numbers, and at least find Cloud Computing important enough to bet large investments in Google App Engine, Microsoft Azure etc on it.

2

The Cloud is lots of things. It basically means you don't care where your computing/storage/hosting is - just that it works as specified and can be accessed on the internet. Because you don't really care where it is, you can move it (whatever it is) anywhere else that offers better service/price/legal status. The supplier typically offers flexibility, expandability and hopefully comaptibility with competitors, and you don't worry at all about hardware or infrastructure issues (beyond assessing whether the supplier can do the job).

So the Cloud could mean:

You pay for Virtual Private Servers to host websites - as you need more processing power, you spawn extra virtual machines. When you need more bandwidth, you migrate all your VPS to a different supplier with lower bandwidth prices or a faster pipe.

or

You pay for 2Tb of storage accessible over the internet via tunneled iSCSI, which is used by your web servers. You then realise that the servers hosting this are in the US, but your data is from EU citizens, so you hastily move the storage to a EU based supplier - if you do it right, your site visitors don't notice.

or

You hire processing time on a botnet to carry out cracking attempts on a banking network. You don't care where the attacks come from as long as they can't be tracked back to you.

There are many other examples that might be called cloud computing. Sure it is hyped and a buzz word, but there is a lot of it going on, and a specific example may be exactly what you need.

3
  • These technologies just needed a generic term which is CLOUD.
    – mhb
    Sep 16, 2010 at 21:40
  • 2
    Yeah - I think it comes from network diagrams where you don't know or care how the connections over the internet are made, so you just put a cloud to show internet and then link the parts of the link you do care about to the cloud. It's vapour, but that doesn't mean it is worthless.
    – dunxd
    Sep 16, 2010 at 22:01
  • Really interesting article on potential cloud meltdown, similar to the banking crisis: blog.gardeviance.org/2010/09/run-on-your-cloud.html
    – dunxd
    Sep 22, 2010 at 12:35
1

I would say it is basically a repackage of many existing technologies, with many improvements to target the mass market.

At the end of the day, all that "cloud" really is, is a scalable, fault tolerant, high uptime, web hosting package.

In most scenarios, it is still cheaper to buy your own hardware and "go it alone" (this website is co-located!), but for many ventures starting up without any backing, it can be a big investment.

To get any sort of fault tolerant webspace can cost hundreds a month (if not more), very few hosts offer this sort of service - usually it is bespoke and they have scripts/people monitoring the environments, setting up additional servers when you need it etc.

With "cloud" all that has happened is that many of the processes in high end hosting have been highly automated to make it cheaper and affordable, but in all essence, you are not really getting anything special that you couldn't get a few years ago if you looked hard enough.

1
  • Great! 10/10 :)
    – mhb
    Sep 16, 2010 at 21:31
1

I have to admit that I'm purposely not going into the "cloud", at least not until I have the time and inclination to do the research required to understand all the implications (ROI, TCO, etc) but it seems to me that cloud computing is somewhat of a throwback to the timesharing days of mainframe computers. Timesharing (even though it's well before my time) was a way for entities (mostly corporations) to lease time on some other entity's mainframe in order to process data, print reports, etc., etc. At some point, computing resources got cheaper and corporations wanted more control over their computing needs and started to build their own in house systems. Several years ago companies (like Amazon) caught on to the idea that they could "lease" or "sell" some of their spare computing power to consumers and voila, a new revenue stream was created and "cloud" computing was born. At this point I'm not convinced that in a decade or two the cycle won't repeat and corporations will bring their systems back in house. What I really think will happen is that corporations will use some mix of in house systems (for those systems where they want fine grained control) and cloud computing for systems they don't want to internally manage (like email).

1

For what it's worth, the "internet" (or really anything once it leaves your building) was always represented as a "cloud" on any diagram I've ever seen. But what I've always gathered about that was it was purposefully drawn that way because people didn't really understand (or want to know) what happened in the cloud.

So to answer your question, yes it's a buzz word. That "cloud" has always existed.

0

It is a billion dollar worth of buzz phrase. So it is the matter who are you asking. I am pretty happy to work in the cloud if they pay me. Phrases don't create any new market. It is demand what does. And if there is demand there is supply as well. This is how economy works. And just because they have to name something somehow it won't be wrong. They call the high-scalability and distributed computing "cloud", live with it. You can call as you want, just make sure everybody know what are you talking about. :)

1
  • So which means in turn to create new opportunities we come up with the re-tweaking of the existing technologies. That's great for innovation. Start something that has already been existed and make it upto the expectation of the market and you are a billionaire :)
    – mhb
    Sep 16, 2010 at 21:34
0

it is buzz phrase in my opinion, it is here since beginning of IT:

  • based on functionality

mainframes with dumb terminals, mainframes with X-terminals, servers with thin clients, lately apps running on the webserver and partially in the browser

  • based on location

offsite/3rd party hosting, in-house hosting still clouds. If you open your network/servers to internet (say through VPN) it becomes also cloud

  • you can choose different level of client devices' "dumbness" (CPU power, storage, ports, etc.)

Probably cloudy or not depends of whether have local storage of data or not.

today there are more business models to buy/sell such services

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .