View RSS Feed


The Cloud (Whether it's iCloud or just normalCloud)

Rate this Entry
by , 12th June 2011 at 03:06 PM (22615 Views)
First blog post, but this is a subject I don't feel comfortable butting into other threads with or a forum thread of it's own, It's entirely opinion and as a result, people are free to reply with their own opinions or ignore it

My thoughts on "The Cloud" have always been fairly negative, and it's highly unlikely my opinion will change on it until the UK's internet infrastructure is 1. up to it and 2. when 1. is able to do away with the home computer as a fat client.

I will and do talk about things like VDI, Thin Clients and The Cloud in the same breath, at least where education is concerned. Internal network infrastructures are good enough to allow a decent level of use in an educational environment, so companies such as The Cutter Project can work their magic and provide a low cost, low power, efficient network setup very competitively. I've always had the same argument against it though; I've watched a school buy in thin clients (iDesks), fill up their server racks with enough terminal servers to offset any savings gained from going fat-thin in the short term and then find themselves having to change their own curriculum. Is that a bad thing? I don't know - I work solely in IT and I'm not entirely aware of how curriculums are organised or why some schools use things like video editing a lot and others do not. If a school has a single suite of 30 PCs, then I could fully imagine a half decent Gb network and TS will handle that rather easily. But spread that out to a larger secondary with 6 or 7 suites of 30 PCs - the network structure would have to be something completely different, effectively a TS/Virtual Host per room, trying to keep traffic local rather than global. But here, I really would need to sit down with someone to understand how and what is possible over such a setup. I won't knock it too much until I fully understand it. I know enough to be able to write it off for my own school.

The Cloud though is a slightly different ball game. It's biggest failure is also it's biggest selling point, which is where I really struggle to understand why it's so well received by so many. It relies entirely on your connection to the outside world. A while ago, when working for a certain ICT support provider, I was enraged to be presented with a piece of software called Junior Librarian.NET. It had the same function as it's predecessor, yet entirely hosted by it's developers. Unreliable? Hell yes - for what was effectively a very basic database hosted off site, I'd never seen so many problems - firewalls, fingerprint recognition problems, even barcode reader problems; why on earth would that be an issue, a barcode reader is no more complicated than the keyboard I'm typing this on. Eventually, the school that was trialling it literally gave up and demanded to roll back to Version 3 of the software, entirely run and hosted by the school. Success. Because the .NET version was too Dependant on the external connection. What if the internet wasn't available - there was no library. That made me laugh a little anyway, because in my eyes the Internet is the biggest library in the world and has already replaced physical libraries in some schools, so this post does have a little irony. So, lessons had to be replanned, people were scribbling book barcodes on paper ready to sort out when the internet was back. Yes, you can have problems with your internal network, but that's entirely at your own control.

Then I was introduced to hosted SIMS - a good old friend run over Citrix. So what happens when internet dies there? No MIS. That's always been a big no-no for me, however that has a slightly two-fold problem. Hosted SIMS was a good and cost effective way to deliver an MIS system to a smaller school. Should that ever be the case? Why should it not be made more cost effective to host it on-site? Or even to a cluster of schools?

This brings me nicely to The Cloud as it exists today. Google Docs. Office 365/Live. iCloud. Online document storage has it's inherent problems anyway - someone I have a huge amount of respect for, Mr Grumbledook Esquire, recently posted concerns about the DPA relating to information stored off-site. If something was to go wrong, let's take an example of a child is kidnapped outside a school gate and the evidence already leads to the kidnapper having prior details about this child. Whether it's Facebook or any other means, the Police and other authorities would need to investigate a huge paperwork trail to gather what information is stored where about this child. If the documents are hosted in The Cloud, could you prove they are secure? Not interested in anyone saying "well, it's a secure site/it has SSL or TLS encryption/the little padlock icon is showing in my browser" because all of that is 100% moot if someone knows a single password. I really need to read up on the DPA properly to say much more about it but I do have severe doubts as it stands. This is all of course something entirely on top of "well if the internet connection is lost, we can't get to that data".
So when the authorities are chasing this information up, they need your cooperation as well as the cooperation of your data host.

I'm not going to say too much about iCloud because everyone is already well aware of my thoughts on Apple as a company. I will say that iCloud is too little too late from them, they were late into "the game" and charging money for what is effectively the same thing that Microsoft/Google/God Knows Who Else offer seems a little out of order/typical Apple. I keep reading blogs and website information everywhere that is touting iCloud as something that will effectively mean the end of localised storage. That's something I can't possibly agree with and I'm not sure the people that subscribe to that theory fully understand what is involved.

"It's an idea that will sell well to all the 'normal' computer users who just use Facebook, Email, the odd Flash game and browse the Internet."

Translated into real terms:

"The average person who doesn't game or do anything powerful with their PC would find this an attractive and possibly cost effective solution, however they won't be aware, or told, what happens when their ISP falls over/Lulzsec or Anonymous get bored one day etc. That SHOULDN'T ever have to be the case but there are more and more security risks coming to light every day, IPv6 is one that needs a lot more looking at as a good example."

So, I am rather anti-cloud. Not in total, though - it does have uses. So I guess I am just very anti-Cloud Reliance. I do not like seeing school curriculums reliant on ICT when it's local (Where's your backup lesson plan?) let alone when the system is entirely out of your hands/control. It has a long way to go yet before that is an achievable goal and a lot of it has nothing to do with the cloud providers - its the UK's broadband/network infrastructure, it's the EU and particularly the UK's naive attitude towards internet security and loads more besides.


  1. zag's Avatar
    Why not get a 2nd Broadband line?

    We have a VDSL backup here and it takes 20 seconds to swap over the gateway if the main one goes down.
  2. znova's Avatar

    interesting post. I've just done a fair amount of research into cloud apps though I am by no means an expert. Cloud seems to divide people into 2 camps like Marmite - you either love it or hate it. I used to be in the former but now I hover around the middle.

    A few points: security seems to be the biggest issue why people are reluctant to take up cloud based apps. However, I would argue that being ISO27001 certified means your security is probably tighter than your average network. But you are absolutely correct, you are still responsible for the data, and most cloud-provider contracts are VERY vague and non-negotiable for all but the largest organisations.

    The second issue is money. Cloud providers generally attract clients with low monthly rates. However, it's like the difference between renting and buying your own house. In the long run (and for some apps less than 2 years) it's cheaper to buy and install in-house.

    I would say at the moment cloud has its uses - but I would stear away from anything cloud-based that is paid for or "business-critical". MIS in the cloud as you said just isn't sustainable with the current infrastructure. However, Google docs etc. could give Microsoft a run for their money.

    I am actually going to be involved in a school cloud project (moving to Google docs, Gmail etc) so I am not really anti-cloud. As I said, as long as it's free, non business-critical and doesn't involve sensitive data(DPA), why not. Obviously the school network infrastructure and broadband connection has to support it!
  3. synaesthesia's Avatar
    Indeed, it's that reliance.
    And as for installing a 2nd line - for anything that's properly useable for that backup, the cost in doing so will instantly wipe out any financial benefits, surely?
    I will stress it appears it's the reliance thing I'm more concerned about, very much along the lines of what you're saying znova. I'm of very similar status, I'm by no means an expert but always keep my ears open for anything that may make our lives easier/better - our being us as technical staff, the schools, staff and pupils. I will absolutely not be happy with anything that comprimises a pupil's education because, along with data protection, that is the most important part of our jobs. Without the pupils, we have no reason to exist in our current state.

    Of course there are benefits, and much of the average techie's job is to weigh up all the pros and cons of everything we do. And those pros and cons will differ for all establishments. I'm also a firm believer that the "one size fits all" ideas are not viable solutions, because it means there are schools and establishments having to comprimise just to conform. I already dislike the one size fits all curriculum but that is something entirely out of my remit or control. Thankfully computers are easier to bunch into categories than people!

    Don't take any of this to heart by the way, for anyone that may take offence at it. I tend to "think out loud" in blogs, that's what I really think it's for. Opinions are opinions of course, and I am very happy to stand corrected!
  4. GrumbleDook's Avatar
    The important thing is not that you think out loud ... but that you think ... and your worries are expressed carefully, considerately and with a decent amount of understanding of the impact that things can have when they go wrong. Keep thinking ...

    I might have to sort out another DPA chat at a #NetworkNorthants meeting again. I'll have a dig around for a speaker of the quality of @Drummer_Boy
  5. SYNACK's Avatar
    Just read this on Slashdot Why businesses move to the cloud: They hate IT | ITworld

    The cloud is a tool so that those who don't actually fully know what they are doing can avoid those that do and make decisions (bad or otherwise). The cloud is Apple for admins, a box of crayons for those that can't work a pen.
  6. Drummer_Boy's Avatar
    The cloud is a hot topic at the moment (I remember when a cloud was an ISP's internal comms infrastructure!), and no matter what our thoughts on it we need to have an answer ready when the bosses come calling!

    I attended the AWS event in London last week (we operate a 'private cloud' environment for our customers, or SaaS as it were 2 years ago!). The AWS vent was very good, but I noticed that the emphasis has subtley shifted to flexibility, and away from cost, as mentioned in a previous post.

    If you need high burst capability, but normally low capacity then the cloud could be the solution for you, as you effectively 'buy' the based level kit, and then rent,m when needed, the rest.

    If you expect to fully utilised a piece of hardware over a period, then running in house is still cheaper - they'll try and muddy the waters with cost of people, but usually this does not stack up in the real world.

    For example, my US colleague is looking at moving to a hosted exchange platform, and the platform providers provided a TCO spreadsheet, based on 5 people working a 24*7 shift pattern, and proving it was cheaper. This was all well and good until I pointed out to my CEO that we only have 2 people working normal office hours, and so the TCO was smoke and mirrors.

    There is a military saying of 'time spent in reconnaissance is rarely wasted', that works equally well for IT. Also, 'Buy once, cry once'.

    PS @GrumbleDook, thanks for the mention.
    Updated 22nd June 2011 at 10:56 AM by Drummer_Boy


Total Trackbacks 0
Trackback URL: