BECTA reports on School MIS and value for money 2010

Quote:

---

Originally Posted by Sdrawkcab

Data redundancy - Wikipedia, the free encyclopedia

Also, that article is misleading. Email isn't really cloud computing in the same way that, for example, citrix or another thin client solution is.

---

Cloud computing refers simply to 'application provision via the web'. The methods for doing that are obviously going to differ, and generally a thin client/citrix solution is not what most think of as cloud computing. Most think of cloud computing as applications moved into a web form and used online via a service such as Amazon EC2, Google Apps, or Microsoft Azure. I've not read of any companies going to a cloud solution for thin client deployment. What would be the value of that? Central hosting isn't cloud computing.

Surely if the data / systems are held offsite, and the service is available in theory from anywhere on the 'net, then it's in the 'cloud'. Perhaps i'm thinking of a looser interpretation of cloud computing.

Quote:

---

Originally Posted by vikpaw

Surely if the data / systems are held offsite, and the service is available in theory from anywhere on the 'net, then it's in the 'cloud'. Perhaps i'm thinking of a looser interpretation of cloud computing.

---

But then, different people have different definitions. The first services to be defined as cloud services were things such as Google Apps and the like. Full thin client provision from central locations has been around for decades - Lloyds TSB have been using it for at least 10 years, for example, and 'the cloud' as a term didn't exist then.

Quote:

---

Originally Posted by Sivadam

......and when the cloud breaks, ............... your data will fall ...........

---

And when your in house server or network goes down? As Ed Whittaker said in the Progresso thread, there's a risk to any data storage in some form or other, so it's a redundant argument. In house or cloud, both still use computers and networks, and there is an argument that data centres are far more resilient than anything a school can afford.

Jees is this cloud vs in house religious war still going? Schools that have in house tech and more money might see the benefits of their own server. School that don't, won't.

I think the solution maybe Azure. I believe you can have it locally or centrally, not sure if you can split it across both.

It ain't that long since we had the Sheffield floods.
Our network was down briefly at the start of one day because of the power cuts.
The web was down or unusable for days!

Cloud obviously finds it hard to mitigate against natural disasters and only in house tech knowledge/servers can help with that. But a cloud that is down a few days every few years still beats the big costs of an in house alternative. Simple economies of scale.

Would the roadworks on the road outside school that cut though our telephone cables be a 'natural disaster' I wonder!

One wonders how big and versatile that cloud will be. Our current experiences with Hosting company that cannot deal with the complexity of our systems are not encouraging. How would God, who controls the clouds, ensure that every single eventuallity was covered? Would all our functionality have to be degraded to a common denominator! Every school would want a slightly different setup!

Maybe they would employ me! Now that would lead to a few arguments! Wot Fun that would be!
Sorry guys I'll be gone by then!

This whole need for 99.9999% uptime for *any* service is a golden target ... but for many services within education that is both unrealistic and unaffordable.

When looking to choose either in-house or hosted (cloud or otherwise) services you need to also do some risk management. Look at both the impact of the risk and the probability. There are plenty of sites and such which can go through risk management with you so rather than roll it all out here again I would suggest that before people start going on about JCBs digging up cables, etc ... they should make sure they have an answer to the question "So you have done a complete risk analysis then and that is why you don't like x?"

How much of this cloud vs in-house is less to with rational argument and more to do with the illusion of retaining control? If you host in house, you feel that you are more in control of your own fate, whereas on the web you are more at the mercy of others. This is the reason why I feel safer driving my car than flying in a plane. In my car I feel I have more control over what happens, over my own destiny; in a plane, you just sit there and let it happen. In the plane you think, what happens if ... (insert you own flying fear here). This attitude of mind tends not to be present in the car because you feel more in control. In reality, this is of course an illusion; I can more mitigate against a drunken driver writing me off than I can against coming down in flames from 50 thousand feet.

I think a proper statistical survey might show that there is little difference in downtime between web and local hosted solution - but then what do I know? I just work here (or is that an illusion too?

Hi,

I see the conversation on cloud and fragility, and wonder what items of essential data would need to be provided in the event that a cloud based system is suddently off-line. As you might have heard during my presentation to the EduGeek Conference, I and my colleagues at Capita think the cloud is probably the future standard delivery mechanism for much of the IT we currently use, but there clearly is a need to provide some essential functions for schools in the event of a connectivity outage.

I assume that a fire register would be one of those items, but would anyone like to pitch in with any other safety critical information that would need to remain accessible that isn't already naturally printed out in the event of a connectivity outage ? For instance an outage of a couple of hourse might not disrupt a school or raise any safety concerns, but in the event of "a JCB pulling up the wires" how much and of what type of data, would need be have a local copy somewhere in the school ? What about if the internet was not available for a couple of days or more; would you simply put in a policy of printing out critical factsheets on children, or would you want a local IT store for this data ?

Just woolgathering here; my personal working assumption would be that schools already print out anything safety or operationally critical just in case the local network goes down, or there is a power outage/disruption, but I've no experience to back up that assumption.

Phillip Hamlyn
Capita.

Hi Phillip

From a Safety aspect it is an absolute must that Registers of some sort have to be available - preferably Session AND Class.
We have a duty of care and must be abale to ascertain the whereabouts of any pupil at any particular time - as far as that is possible.

It is totally impossible to maintain 50 Reg Group Registers and a myriad of Class Groups Registers without access to the system. Printing Paper Registers, 'just in case', every time there is a Timetable or Group Membership Change is totally impracticable!

A record of Cover is also vital as it is another absolute must that a member of staff is in front of every class at all times! (Chris Sherwood will have heard that one before from me quite a few times!). Fortunately we can save to HTML and print a Cover Summary in advance (i.e. tomorrow's!). We now do this as a matter of course but we only print a paper copy one day in advance!

Update
There are loads of other areas too but I guess these will differ for different schools.
For example, I still maintain NT4 even though I have been using NT6 since it's inception. Each time I make a change to the timetable I take a copy of the updated NT4 Timetable and load it on to my Datastick. NT4 runs as a satellite installation directly from a datastick, so I always have a timetable available even if the electricity goes down (as long as I can find a laptop with some juice in the battery!)

Basically, i think most would like it to work by caching all data locally, whilst having the cloud server as the main home of it all. So, even if it went down for a couple of days, the local school would be able to carry on as normal using the local cache.

Limiting the access down to specific things wouldn't work as the list is pretty long. Medical data, attendance, contact details, timetables, behaviour, etc... all get used every day, all day.

Sivadm,

I think I understand. So a key component of cloud based solutions, in your view, would be the ability to have a locally accessible set of data, and an application (hopefully the same one as works online) to access it ? I can see the need for class registers, but presumably these would be "incomplete"; i.e. an indication of where the child should be, not the actual fully functional register itself ? Presumably you'd need a child emergency contact sheet as well for each pupil, or would that normally be printed off and kept in a cabinet somewhere ? I'm trying here to get a feel for the scope of locally cached data rather than for specifics.

Phillip Hamlyn
Capita.