Presently I have to determine whether to read more of the blog, or go to bed….

Decrease your IT costs, my ass. ISO XXXX certified? With a datacenter you don’t know? Please.
How is your data being redundantly stored in the cloud? Is it stored, redundantly, in the same server rack as the primary? Where, physically, is your cloud datacenter? Is it as safe as a NORAD bunker, or is it in a trailer in tornado alley? Does it have any security at all? You’d better hope so. The IT staff you’re booting to ‘reduce your it costs’ are certainly going to test it’s integrity.

I have my fingers crossed hoping for a massive ‘storm’ to lose everyone’s google docs permanently. If you’re debating using cloud computing, at least do yourself a favor and get a tour of the server farm, ask questions, and plenty of them, and also realize that this company you’re talking to is NOT going to take responsibility in the event of some catastrophic data loss. Sure, they’ll SAY they’re going to, but I bet they spend as much on lawyers as they do on security.

Signed,
The dinosaur on his mobile workstation.

I.E = Internet Explorer is a software , so a software will control internet explorer? hmm that would be windows

Utility computing was the vision, cloud computing is the reality.

There are no standards (too early yet) except the general agreement that a REST API is preferable to SOAP (sanity rules!).

Interestingly the Microsoft storage API is much richer than Amazon’s and may offer MS a significant advantage when they finally get their pricing sorted out.

Thanks for some sensible replies. These are the definitions of Cloud Computing that do make sense but are drowned out by the noise. I really enjoyed the Joe Weinman articles which should go straight onto the Wikipedia page. I like the way he says that Cloud Computing is orthogonal to virtualisation – that makes sense. I shouldn’t care if the cloud is physical or virtual. I also really enjoyed these articles: http://www.the-network-effect.com/

I still don’t understand if there is a major difference between this and what we used to call Grid or Utility computing. Or is the idea that with Grid computing that there was only going to be one Uber-Grid, while we’ll have a lot of Clouds run by different companies?

And what exactly are the standards that the different working groups trying to address? Are they aimed at Cloud Builders to ensure they build them in a standard way? Or aimed at an interface which Application Builders can use to ensure that their apps can be hosted on any Cloud (maybe a bit like a language independent OSGi)? It feels to me that the standards are going to be more at the infrastructure level than the application level, which is maybe why people like me that play in the software tier don’t completely get it.

There is no doubt that these powerful clouds, near infinite resources and different pricing models are going to change things. But they already have. That’s what *aaS was about. Is there any reason to expect that the Cloud-* standards are going to arrive any more quickly than the much hyped Grid-* ones?

1. Virtualised Infrastructure (i.e. controllable solely by software)
2. “Pay as you go/pay as you use” cost structure
3. The ability to scale down as fast and as easily as you scale up
4. No cap on resource utilisation

Nick Carr and are Joe Weinman two of the better folks with explanations of this phenomena and I think one key is to separate cloud and virtualisation. One way to build a cloud is virtualisation but they should be treated as separate. in the same way that SaaS can be delivered from a cloud computing platform but it’s not essential.

for me, cloud is more defined by things like consumption based computing or utility computing than the technologies used to deliver it.

have a read of http://gigaom.com/2009/04/11/6-half-truths-about-the-cloud/ and http://gigaom.com/2009/04/19/another-half-dozen-half-truths-of-the-cloud/

plenty of good stuff in there to sort the wheat from the chaff

Steve

Your rant rocks. Strongly recommend it to everyone. Our sentiments are almost identical. The main difference is that you seem to prefer being shot with a nailgun, while I’d rather take a normal bullet and end it quickly. And you understand Utility Computing better than I do.

Blogroll += Shane Johnson

“I purchased a cloud. Did you? Wait. What is a could?”

http://blogs.citytechinc.com/sjohnson/?p=84

Shane

I like your buzzword list. I’d put “SaaS, Web 2.0, Enterprise 2.0, Social Web” into the Marketing Buzzword Bucket. They haven’t helped anyone. I think AJAX was useful as it named something that needed a name. Reminds me of Ursula Le Guin’s Earthsea and “the power of true names”. Web Services always had focus and substance but because it had a catchy name instead of something like “Discoverable-HTTP-XML-RPC” the marketeers grabbed it and confused people for a while.

I think that “Grid Computing” is the closest thing to “Cloud Computing” by a long way. It’s nothing to do with the Internet, really. It’s about a new level of abstraction that we need to understand super-powerful, self-managing, MapReduce enabling virtual machines. And about the social-economic impacts that come with this – privacy, licensing and a whole lot more. They might even “solve” Chess by analysing the entire game tree to decide that White Wins or White Draws. That’ll be weird.

I’d go as far as to say that Wikipedia would benefit if someone took this http://en.wikipedia.org/wiki/Grid_computing, did a little s/Grid/Cloud/g, and updated their Cloud Computing entry.

The Cloud is undoubtedly a buzzword for application service provision, but I don’t think we should knock it. If you say it’s just the internet, while that’s true you have to remember that most people don’t know what the internet is.

The other aspect is that players with huge resources like Amazon will be able to provide storage and power at a cost and flexibility that traditional co-lo and managed service providers cannot; that virtualisation will be increasingly important; and that this has a direct impact on hardware provision.

This is why Larry might be so tetchy about it: he’s just paid for a whole load of kit in order to get an application server. Moreover, his biggest competitors in Microsoft and Google and going more and more cloudy. So you’d expect him to berate them.

If the Cloud stops people talking their usual drivel about enclosed enterprise applications and gets them to understand the internet’s potential better, then I think as web development professionals we should be happy about it.

Let’s review: Cloud, “Grid Computing,” SaaS, Web 2.0, Enterprise 2.0, AJAX, SOA, Web Services, Social Web … all buzzwords that tried to go too far, do too much, while wrapping things that already existed under better-understood names. Out of these, Web Services has the most concrete basis (as buzzwords go); it actually refers to something. But even WS was quickly hijacked and its meaning perverted by pundits and jackass analyst-types.

Er… hey wait a minute. I *am* one of those analyst-types…

The first item on the Open Cloud Manifesto FAQ is:

Q: What is the Open Cloud Manifesto?
A: The Open Cloud Manifesto establishes a core set of principles to
ensure that organizations will have freedom of choice, flexibility, and
openness as they take advantage of cloud computing. While cloud
computing has the potential to have a positive impact on
organizations, there is also potential for lock-in and lost flexibility if
appropriate open standards are not identified and adopted.
Key members of the cloud community worked together to produce this
document and endorse it to establish a set of core principles around
the open cloud.

I’d like to see a question before that. What is Cloud Computing? Cause I couldn’t answer that right now. My favourite answer so far might be this one, which has very little to do with all the hype: http://cloudcomputing.sys-con.com/node/579826

Personally I have a long standing respect for ‘Big Blue’ though and its deep understanding of knowledge and information management. Know doubt there are more informed technical opinions on this but experiences in helping implement its Quickplace/Workplace solutions pre SharePoint demonstrated that its thinking on business collaboration was way ahead. Granted SharePoint popularised and gave ‘collaboration’ mass business market appeal but I have rarely had a SharePoint implementation experience from a user perspestive that was as good as the IBM capabilities that pre-dated it.

So – when IBM starts making rumblings around things like the Open Cloud Manifesto I start to think that a descriptive concept, which as you rightly say has little substance, actually has some deeper, real substance underneath.

I had the same feelings about SaaS over ASP. Although I had the same feelings about AJAX as we’d all been doing asynchronous JavaScript calls to get XML before the catchy name came along. It is true that naming a technology and generating a bit of hype around it can actually drive the standards and toolkits. But as we know, real clouds have very little substance to them, and this buzzword has even less.

I hadn’t seen the article you mention. Interesting read. Not quite sure I understand it properly yet.

Back in those ancient Web 1.0 days (or possibly Web 0.5) any techie meeting I ever sat in had ‘The Internet’ represented by and described as ‘The Cloud’ and as you so rightly say the hype has blown out of all proportion.

When I was working in the hosted services industry a few years back I got similarly disillusioned with the hype around ‘SaaS’ as I couldn’t really see that much difference between what was being described as SaaS and the Application Service Provider model back in the late 90s – other than the more widespread availability of higher bandwidth and therefore a greater uptake.

Did wonder whether the following has some relevance here in IBM’s recent moves to agree some standards and also its view on what sits in The Cloud … http://www.readwriteweb.com/archives/ibm_blue_spruce_first_look.php

And as for your last paragraph – I’ve finally discovered what it is I’ve been doing wrong all these years