Mention cloud computing to a business hack and you’re more likely to be given a restraining order than an interview.

“Not cloud computing,” they’ll groan. “But while you’re on the phone, does your client have a life-threatening illness and live in the North-West?”

cloud-vomit

Or something like that. But apart from what’s perceived as massive over-hype, cloud computing’s got real problems. A lot of people don’t know what it means. Even more think it somehow sprang up overnight to save small businesses everywhere, when the reality is that it’s a term describing manifold and incremental changes across decades in how we use IT. The web’s awash with IT companies propagating obfuscating Aascronyms.  But here are the kickers for business – can you feel the bile rising?

Security
It’s pretty clear that companies aren’t going to start whacking up lumping great chunks of mission-critical infrastructure into the “cloud” en masse. They just aren’t going to do it. There are no universal standards in place. And the all-encompassing term “cloud” hides the fact that data is actually stored right here, on Earth – and as Jamie Turner said recently, “you can have as much redundancy as you like at any given site but if it disappears into the San Andreas Fault you’ll be wishing you still had that magic DAT tape.”

The only reason CRM (e.g. Netsuite and Salesforce), for instance, has succeeded in the cloud is because there’s no viable alternative. This has forced people into trying it out. They seem to be happy but nevertheless there has to be a driver for people to take the plunge elsewhere.

Privacy
I’ve already discussed “Twittergate” here.
This is linked to cloud security and there are no concrete answers here either.

Control
This is my favourite. The irony is that often, if you outsource some function to the cloud, the guys who look after that aspect of your business operations actually do a far better job.

They should be specialists in that area, after all. Uptime, for example, is probably considerably better. The only problem, of course, is that when you’re totally in control of your own infrastructure you don’t mind so much about hiccups… you might even discreetly write them off. You might grumble but you won’t start tearing your hair out if something has to go down for a while.

Five minutes of down-time for a cloud-based operation can spell disaster and a reputation in tatters. Businesses need to choose suppliers with a good track record but they also need to accept their lack of control or make sure the service fails over elsewhere in the event of downtime.

Conclusion
“Cloud computing” covers a range of IT practices that really can benefit business. But there are problems, and they need to be addressed intelligently. “Cloud computing” – a dumbed-down blanket term which covers far too many processes – is a phrase which more often than not just gets in the way.

Advertisements

The latest Twitter controversy surrounding the blog, the hacker and the cloud vendor isn’t disturbing – just inevitable. By now anybody with an iota of interest in cloud computing will know what this story is about. Many people are probably damning Google for their ” lack of security.” But hang on here. Aren’t people being quite cavalier with their data? The other day I refused to give my own partner my PIN… but as I write, it’s happily stored somewhere as a draft on GMail. That really doesn’t make sense.

Hell, I trust the cloud more than I trust myself

Who’s really to blame? I don’t think it’s black-and-white. Frankly, as a rule I trust some company I know nothing about  a  lot more than I trust myself. I leave my passwords lying around on the desktop. I write my PIN on a scrap of paper and keep it in my wallet next to my debit card (nobody’s fooled by the fact I’ve made it look like a phone number). I’m lazy and useless – and I suspect most people out there are too. However, I think cloud vendors have a responsibility to make sure they compensate for users’ inadequacies.

Keeping sensitive data in the cloud isn’t “probably going to happen” – for consumers, it’s been happening for years – the big vendors just need to pull their finger out. At the moment, if you get stung by a lack of cloud security you’ll just be told:

“Only a dribbling buffoon leaves all their valuable data in the cloud.”

While it’s true that simple passwords were used – and in this respect Google is relatively blameless – there really ought to be more safeguards in place so people are forced to at least set more secure passwords. This is a must if the  business cloud is going to expand from web services and utilities into other areas such as secure data hosting.

Jamie Turner, UK cloud computing evangalist and IT Director of TheWebService, has this to say:

Cloud storage, just like Esperanto and the Sinclair C5, is a concept that makes sense… almost.

Having access to your data anywhere in the world from any device is an incredibly powerful thing. The scope is huge, enabling much wider usage and new possibilities, most of which haven’t even been thought of yet.

The problem is, making your calendar universally available in the cloud is a very different thing to placing business-critical company and customer information out there – especially financial information. Despite the significant business drivers that may promote this approach – scalability, agility and all the things that basically remove the inertia that blights most IT departments – security’s still the show-stopping concern.

There are too many questions, and too few answers. What control do we really have over data once it’s up there? What’s the physical security of the data centre? Where is the data centre? Are there cross-border legal issues with hosting the data overseas or in territories with ‘incompatible’ legislative environments? What if you need to destroy data – is that even possible? Then we need to consider the availability of the data: what if the cloud provider folds or they’re taken over by an overseas organisation? If there’s a catastrophic data centre failure, what’s the recovery time? Do they even back things up or just hope that a single data centre will always be safe? It’s a glib but important question – you can have as much redundancy as you like at any given site but if it disappears into the San Andreas Fault you’ll be wishing you still had that magic DAT tape. Are we blindly throwing data into the sky in the hope it will stay safe? Ultimately, this is the big problem with storing sensitive data in the cloud – at least for now: there are just no convincing answers to any of these questions.

So let’s not throw the baby out with the bathwater – this problem just needs to be addressed, fast.

Microsoft took their time, but the web is abuzz with news that Office is going to be available as a fully web-based service.  According to ReadWriteWeb:

These will be “lightweight versions”, but Microsoft told us yesterday that they’ll still have rich functionality and will be comparable to Google’s suite of online office applications. The apps will enable users to create, edit and collaborate on Microsoft Office documents through the browser. The apps will work in IE, Firefox and Safari browsers (no word on whether Google Chrome will be supported).  Microsoft clarified in an email that these apps will use HTML and AJAX, but also Silverlight components.

The next release of Microsoft Office will include browser-based versions of Word, Excel, PowerPoint and OneNote.

E-marketers are missing an opportunity to tell the world how eco-friendly they are

In the SOA industry – infact, any industry that operates at all “in the cloud”, being “green” is a differentiator that can swing business your way. In this digital world where price plays such a huge part in purchasing habits – whether B2C or B2B – it actually gives you an opportunity to up-sell your SaaS services.

While observers in other sectors predict eco-friendliness becoming so ubiquitous it ceases to be a viable USP, the digital world has remained remarkably resistant to being more sustainable.

I caught up with Guy Mucklow, managing director of UK-based web services company Postcode Anywhere, who I (obviously) work closely with and who is an advocate of being green and the green marketing that goes with it: “This shouldn’t be about ‘greenwashing’ – the deplorably cynical business of claiming everything you do is green, indiscriminately, in order to win a few customers cheaply,” he tells me, taking a break from typing.

“Most digital businesses are naturally going to be greener than their traditional counterparts: if we consider e-commerce, for instance, there are less vehicle miles, lower inventory requirements (pre-selling means lower production of obsolete goods), less printed materials, less packaging and altogether less waste made than with traditional businesses.

“The fuel emissions saved by switching to online grocery shopping is estimated at between 18-87%[1] – pretty impressive! So why not tell everyone about it? Because there’s a problem… it’s just not quite enough to say ‘we operate online, we’re green, buy from us.’ It’s pretty lame and certainly not a differentiator if your competition is mostly other companies operating in cyberspace.

“But it’s pretty easy to go that extra distance. Look at other organisations, especially (dare I say it), the public sector. What about carbon-offsetting your website at www.carbon-neutral-website.org ? And if you’re not sold on the benefits of carbon-offsetting, why not switch to a green energy provider?

“Remember: no greenwash! These changes can’t happen overnight and it’s a sensible idea to spend a day drafting up an eco-plan listing all the areas that can be improved and imposing a sensible timescale on achieving them.

“That’s what we’re doing at Postcode Anywhere. We’re already virtually paperless and more changes are on the way. And it fits in with our corporate culture – one of our latest services is a route optimiser that can cut haulage companies’ petrol usage by 30%.

“In fact, excessive fuel consumption is one of the stumbling blocks for e-commerce’s green appeal – if there’s a tangible product to be delivered, they mess it up on the last leg by plotting lousy delivery routes.

“If all the small-medium haulage businesses (who generally aren’t as green as the big players) used our route optimisation, we would see a 13 million ton drop in CO2 emissions every year – more than twice the yearly emissions of Teesside power plant.

“So that’s something to think about too. If you run a large e-commerce company you might want to consider taking the plunge and going with an established software company that offers route optimisation like FleetRoute. For the smaller companies, we have developed a pay-per-use product at Postcode Anywhere which typically costs a few pence per use – opening up route optimisation to everyone. It’s green and it will save you serious money as well.

“There’s so much you can do. You can get free water saving save-a-flush bags from Severn Trent (stick them in your cistern and the job’s done!) and you might be able to get free energy efficient light bulbs from Warwickshire Energy Efficient Advice Centre (WEEAC ) – visit their website at www.weeac.org for advice. And what about working from home where possible? It saves petrol and is very viable for web-based businesses.

“In short, making genuine changes to the way your digital business operates should be a lot easier than it is in other sectors. You’ll help the environment. You’ll offer a more attractive product or service. Going green is a no-brainer. All you need to do then is tell people about it!”

Information on Postcode Anywhere’s affordable route optimisation software is available at their website.


[1] Source: Siikavirta, H, Punakivi, M., Karkkainen, M and Linnanen, L. (2005) Effects of E-Commerce on Greenhouse Gas Emissions, A Case Study of Grocery Home Delivery in Finland. Journal of Industrial Ecology, Vol 6, No 2, 83-97.

Web services, SOA, everything-as-a-service: what are all these terms?  What do they mean to you?  Here’s a list of definitions as I see them:

SOA
Service Oriented Architecture.  a way of describing a business model that uses cloud computing.

Cloud Computing
The use of Web Services.

Web Services
Software or data served over the Internet as an ongoing service rather than a custom-made or out-of-the-box piece of software or dataset (delivered, say, on CD)

DaaS
Data as a Service.  Outsourcing data managment to “the cloud” – either “public” data (such as street names) or private data (such as stock lists).

Desktop as a Service: This term hasn’t really caught on due to DaaS being generally reserved for “data as a service” or “database as a service”.  Desktop as a service is a rather more accurate term for the more buzzword-y “Operating System as a Service”; not a true operating system (which boots up the computer and manages a user’s hardware resources) but an online desktop which stores your preferences and basically acts as an interface for other web services, just like a normal desktop is a user intterface for traditional software and data.

SaaS
Software as a Service.  Outsourcing software to a company “in the cloud” that runs the software, updates it, and allows business access to it for a charge.

PaaS
Platform as a Service.  A piece of software delivered as a service that allows development of 3rd-party apps to “plug into” it.  Examples: Salesforce.com, Facebook.com

Lesser-known services under the umbrella “everything as a service”

CaaS
Communication as a Service.  Delivering telecommunications, instant messaging etc. as a service over the Internet.  Telephony as a service, also known as “Voice as a service”, employs VOIP (Voice Over Internet Protocol).  Software and hardware can be provided as a service by providers.

IaaS
Infrastructure as a Service.  Slightly tricksy one, this.  Can be taken as being similar to “Operating system as a Service” or “Desktop as a Service”, Infrastructure as a Service typically dishes up services centred around hardware as well as software.  Wikipedia tells us that “servers, software, data center space or network equipment” are all served in IaaS.

MaaS
Monitoring as a Service.  Network/website monitoring delivered as a service.  Notifications of network or website problems delivered via SMS, email, etc.

Another term not really part of the Web 2.0 Buzzword Boom but still applicable

Utility Computing
Self-explanatory: providing computing services as a metered utility, in the same way that the services above are delivered as a service.

I know I’ve missed some and will update this list and make it more accurate on a rolling schedule.  This can be another DaaS: Definitions as a Service – not that I’d want to muddy the waters any more, you understand.

You can read here for a bit of talk about SOAP etc.

What is cloud computing?

This isn’t another quick-start guide to cloud computing and Service-Oriented Architecture (though you can follow that link to get to one); it’s an investigation into what people are searching for on Google.  So what are the top terms for Cloud computing?  It’s no surprise that “what is cloud computing?” is near the top of the list.  Braving Google’s wrath by my apparent keyword spamming, here they are:

cloud computing
ibm cloud computing
what is cloud computing
microsoft cloud computing
cloud computing wiki/wikipedia
elastic computing cloud
cloud based computing
cloud computing definition
cloud computing services
cloud computing market
cloud computing security
computing in the cloud
cloud computing conference
cloud computing software
define cloud computing
gartner cloud computing
sun cloud computing
cloud computing blog
cloud computing solution
free cloud computing
cloud computing uk

So what does this tell us about people’s relationship to cloud computing?  Well, you can’t really tell unless you know how many people are performing the searches.  Surprisingly, Google tells us that while 27,100 people per month searched for “cloud computing”, only 480 per month asked the question “what is cloud computing?” … and from there, the numbers trail off quickly to just a trickle of people making searches on the terms lower down.

So does this mean that people are really quite savvy with the concept of SOA?  I don’t think so.  I think the 27,100 figure is significant, and that many of the people performing this most basic search are in fact looking for a definition of cloud computing.  The fact that the search pattern is so grossly top-heavy, with not so much a long tail of minor searches as a long dribble tells me that people either have a vague idea of the term, or they’ve never heard of it at all.

Is this surprising?  Perhaps it should be, because cloud computing has been around for longer than a lot of people think.  (And I am talking about the current set of “buzzwords”, not reflecting on how “computing’s come full circle” with these new but old concepts of dumb terminals etc.)

Perhaps the main problem is that there are just too many terms flying about.  Let’s try and categorize them all:

Cloud Computing
“in the cloud”
Service Oriented Architecture (SOA)
Web Services
Software as a Service (SaaS)
Platform as a Service (PaaS)
Data as a Service (DaaS)
SOAP
WSDL (or even wisdl)

I think these are the core components of the terminology attached to SOA and cloud computing.  Actually, looking at them, they don’t look too bad.  That’s when they’ve been roughly sorted into some kind of conceptual order though.  Mix ’em up and introduce them to the uninitiated and it starts to look worse.  The biggest problem, though, is that nobody’s terribly sure what all of these acronyms and even backronyms are supposed to stand for.  Take DaaS – it could mean any of:

Data as a Service
Database as a Service
Data Warehousing as a Service
Desktop as a Service

It’s also a Belgian beer.  Now this uncertainty of what all these terms are actually supposed to mean is not exactly helping people get things straight in their minds.  And worse, the terms all interlink in one horrible mess.  I’ve tried to straighten things out a little with this diagram:

Web services, SOA, DaaS, SaaS and PaaS

Cloud computing diagram: Web services, SOA, DaaS, SaaS and PaaS

But really it’s the opposing forces of everyone trying to peddle their own definitions that causes many of these problems in the first place.  Not that I’m going to desist… of course, my terms are the correct ones.  I mean, let’s look at the tangled web of saying DaaS means “desktop as a service”: It’s a desktop as a service.  Which makes it also a platform as a service.  It’s also a kind of operating system as a service (I guess they’re trying to appease the pedants by veering away from calling it an OS as a service) as well as a manner of software as a service.  Hang on… it’s everything!!  It can fit into all the categories.  These terms slide over each other like Venn diagrams of doom, tectonic plates set to annihilate each other with obfuscation.

So let’s just stick with the diagram for the terms SaaS DaaS and PaaS and call it a desktop delivered as a web service.  That’s not confusing.  Order is restored.

Tomorrow I might take a little look at some other popular search terms and try and see where they fit into the picture.  I’ll also be establishing a comprehensive lexicon that is transparent and easy for everyone to get a handle on.  In the meantime, I would be interested to hear about people’s definitions of the various acronyms flying around out there.

Linux is the answer if cloud computing & Service Oriented Architecture (SOA) is going to work… this open-source operating system is the future of SOA & cloud computing.  Let’s present the argument:

Microsoft knows what it’s doing with SOA and Vista

Traditional operating systems will be dead in a few years.   they’ll be killed by service-oriented architecture.  That’s what my money’s on and if you think I’m a rabid Web 2.0 nutcase then you can’t see the bigger picture.  Microsoft have realised it.  In fact, that may (somewhat ironically) be one of the reasons Vista was such a retrograde step in opening up Web 2.0: Microsoft wants to milk us for as much as possible, and the best short-term plan is to positively hold back cloud computing & SOA for as long as possible.  They know Service Oriented Architecture is coming, but as the biggest player by far in the exisiting OS market, why accelerate its growth?  It makes sense to stave it off as long as possible and rely on a business model that they know works.

The reason Microsoft is starting to come round now is that while they don’t wish to accelerate the advance of cloud computing (and miss out on flogging us Windows 7, 8, 9 & 10 for our very earth-based desktops) they want to be the first to the table once a certain tipping point has reached and cloud computing is not only inevitable but immediately forseeable.

Imagine a dog.  This dog has knocked a big butcher’s block off the kitchen table, and two big hunks of steak have fallen off.  One steak is within reach, so he sits there and starts work on it.  The other steak has landed on the other side of the room.  Now, he’s not the only dog in the house.  There’s another one upstairs.  This dog upstairs has heard the commotion and starts coming down.  Now does the first dog immediately rush off to defend the second steak before the first dog gets there?  No.  He’s got enough time, so he wolfs down the steak he’s got before doing an about-turn and pouncing on the second steak, before the second dog gets there.

OK… this is a very laboured and imperfect analogy: the first steak is the existing operating systems and software market, the second the cloud computing and web services market.  The first dog is Microsoft and the second represents its competitors.

The Future of SOA, SaaS DaaS and PaaS

As I forecast in this post, one sensible (but shocking) strategy for Microsoft is to “leap-frog” the market and invest heavily in cloud-based “operating systems” platforms and use this “operating system” to roll out software-as-a-service.  There will be a place for wholly deskbound solutions (late adopters – mostly home users) but this market will be a fraction of what it is now.  I predict Microsoft will concentrate their development on the platform in the cloud and stop pushing their desktop operating systems so hard.  Why?  Because more and more people will be running Linux.  It’s free, and it’s becoming more and more widely supported.  Why should people invest in two operating systems? They won’t.  They’ll use Linux to boot up their computer and use a cloud-based system to interface with software, their desktop, etc.  But it’s going to be messy.  Some people will be running Linux, some Apple, and some will be using Microsoft’s latest home OS.  But the real future is in the cloud and whoever can create the most ubiquitous, one-size-fits-all solution will dominate the market.  Remember, we’ve already seen it with Microsoft and Windows: by leaving the hardware to others (which was seen as a ghastly mistake at the time), Microsoft made an awesome coup in the OS and thus software market.  It scored twice by capturing a platform: it sold the OS and could roll out the software too.  Now Microsoft, if they have any sense, will concentrate less on the earth-bound operating systems market for everyone to fight over and concentrate on making a solution in the cloud.  It should stop beating us round the face with more bulky additions, millions of add-ons and lines of code in its OSes and give us a neat, trimmed-down, lean, super-fast desktop OS that hooks into the real delivery system – the OS in the cloud.  But the OS in the cloud is where the real money will be made because more and more people will switch to Linux – and eventually Microsoft’s earth-based OS business will dry up.

Linux users need to get their heads out of their asses

There’s something about Unix users.  They’re often too clever for their own good.  They’re often smug.  They’re always bloody clever.  They’re a part of a club and often like to look down on the people who use Windows: the double-clickers, the Office buyers, the Wizard users.  Unix users like to have a relationship with their computer.  It’s more personal.  They’re probably built their computer themselves.  I have a suspicion that many Linux users are pleased that it’s more troublesome to use for the average user.  It keeps the riff-raff out of the club.

Here’s the opinion of one Linux user, Scott:

I am sorry, but the Linux world has got to get over the “geekie-ness” and get something out that is for the users. Yes, you can build it, customize it, make your own distro for all it’s worth. The bottom line is that people want an OS that runs the software and hardware they use.

I am an avid Linux and long time Mac user and I have to say, I feel like I am fighting with the OS much of the time. Getting drivers to work is just one of my pet peeves. Yes I can spend my time figuring it out on google, but why? Say what you want, but I can boot up OS X or Windows and they all work (software and/or hardware)…..and you do not need to be a “geek” to get them to work. OS X is what Linux should aspire to be….simple, powerful, easy to use, with enough play under the hood to satisfy any geek.

We can say…virus free, runs on old hardware, and “I do not bow to the MS empire” all you want…but the bottom line is that Linux does not work or run the software or hardware people want….in the home or business. If it did, do you not think more people would be using it?

What I do find funny is that the Linux/OpenSource community may have bigger ego’s and heads that Mac users.

Well, Linux users may take umbrage at my sweeping generalizations, and some may argue that it’s not in their interests to become mainstream anyway (they’re happy as they are) but as I see it, a chance for Linux to come into the mainstream is a major opportunity for a great OS – after all, what’s irritating for existing users of Linux at the moment?  Lack of support from hardware manufacturers, that’s what.  Getting drivers that work.  It’s not mainstream so it’s not supported, so users have to pointlessly wait around for the community to hack the problem.  Adoption by the mainstream would eliminate this problem.  The community would grow.  Linux would be the ultimate King of the operating System.  It’s free so it would become pre-installed on all systems.  People who wanted to “upgrade” to Windows would become fewer and fewer.

Windows users need to get their heads out of their asses

There is a great culpability attached to Windows.  OK, things might go wrong, but you’ve always got someone to blame.  In fact, it’s become almost a sport.  It doesn’t really matter if they can solve your problem or not (which is, I suppose, a good job); the point is it’s Not Your Fault.  With Unix, you have no-one to blame but yourself for choosing the bloody stupid free stuff in the first place.  In this sense, there is a problem.  As I see it, the solution is one of these:

a) Retailers offer Linux customer support (but they won’t know what the hell they’re talking about)
b) Hardware manufacturers offer customer support (Massively unlikely)
c) Businesses are set up that offer a flashy front-end to Unix and charge a nominal fee… mostly for customer support

Asking the community is not enough for the late majority onward: they want a number to ring.  A figure head to hate.  Someone to Blame.  They’ve never been sure of computers and they pretty much need that.

Service Oriented Architecture: Exisiting Open-Source Solutions

Take a look here for some discussion on gOS 3.

Advances are being made and for me, the future of SOA, software, data and platforms as a service is very clear.  As ever, I heartily recommend taking a look at The Web Service website to have a look at some of the possibilities of Data as a Service and cloud computing.

As ever we’ll continue to look at Service-Oriented Architecture: Software as a Service, Data as a Service and Platform as a Service on this blog… if only to ask the pernenial question: “What is Service Oriented Architecture?!”