Cloud Computing and Management Information Systems- what do members think?

[Mike_Bostock]

I thought I would add another contribution to this thread having attended the National College Conference exhibition this week. Whilst there I tried to remember to ask the Cloud question of headteachers who visited our stand. Looking back, it was probably only about 10 heads and another exhibitor who I asked about Cloud MIS, but reactions from this small sample were interesting, nevertheless.

 

I can report that only one headteacher I spoke with had a specific view on this question, and that head made the point that their VLE was Cloud-based and it would be more convenient for connectivity if their MIS was as well. For others their prime consideration was that their MIS worked reliably and they weren't particularly bothered about where the application lived.

 

My interest in this question lies in the fact that the system that we market draws data from different school MIS solutions. At the present time, over 90% (probably closer to 95%) of MIS in schools are installable, networked, SQL-database driven Microsoft applications. For our system to have a rich connection to the data fields in an MIS there is no question that it is far more efficient because our application is also an installable, networked, SQL-database driven Microsoft application.

 

The main point made in my discussion with a supplier who markets a related Cloud-based product was access from home by teachers. Thinking this over, I am still of a view that if it was me, I would prefer a VPN solution that also gave me access to all my other files and applications on my work computer.

 

So, for me anyway, I think I will wait a while longer until the percentage of MIS in schools that are Cloud-based grows a bit, before putting resources into developing a Cloud version of our application.

[plus4schools]

Thinking this over, I am still of a view that if it was me, I would prefer a VPN solution that also gave me access to all my other files and applications on my work computer.

 

Good points.

 

On VPN:

VPN can be a halfway house solution to offsite access without going completely cloud, but the major problem apart from added complexity is that it’s unlikely to be scalable or very responsive in most situations because the outbound bandwidth on low cost ADSL lines is very limited.

 

That’s not normally a problem for web surfing etc where the inbound data from websites far exceeds the few characters of outbound data you type as the URL to access it – the very reason why ubiquitous broadband was designed to give asymmetric bandwidth, which as we know is what the A stands for in ADSL. If you start serving content up outbound from the school either using a local web server or a VPN connection to an MIS server located in the school, the capacity requirement is reversed so you are in effect swimming against the tide. That’s bound to place a limit on the number of simultaneous users and/or severe reductions in response times.

 

You also have to think that usage of school systems like report writers, parent portals and MIS in general etc is very lumpy because everyone tends to do the same things at more or less the same time and statistical bandwidth levelling that you find in other non-school situations doesn’t work.

 

Furthermore, VPN can often be a bit problematical on home broadband connections for other reasons such as traffic shaping limits imposed by ISPs. Depending on the ISP and package you have, the actual throughput of data on VPN ports may be slowed as they prioritise other things like web surfing to balance the load on their backhaul internet connection at busy times e.g. evenings and weekends (i.e. non school hours). Most of the big name ISPs don’t even tell you that they are doing this and those that are honest about it usually offer a ‘home worker’ or a similar package that does give a higher priority to things like VPN at a higher cost. People in this forum can probably figure out the best way through the stuff ISPs don’t make obvious, but the average staff member? Even if they do understand the argument, I doubt many would change their ISP because of it.

 

There’s also the security aspect of using VPNs to consider – if that bridge is breeched, then intruders have access to a lot more of your data than just the MIS, so you need to be really careful about how you set it up.

 

You will also find that many low cost broadband routers (virtually anything sub £500) will have a fairly small practical limit to the number of simultaneous VPN tunnels that they can create due to the limited power of the internal processor, especially if you use a decent security policy. There’s always DSL lines and Cisco of course, but not many schools want to spend that sort of money...

 

On the other hand, cloud hosting is designed to be scalable and provide reliable connectivity and you don’t need anything more than a basic cheap as chips broadband connection at the user end wherever that is, at school or at home. That also makes the case for web based MIS systems vs. standard apps running on a remote cloud server, because ports 80 and 443 used by web servers will always get high priority from any broadband traffic shaping.

[PEO]

As it stands now, I think cloud computing is still relatively new and the broadband in this country is p155 poor. Give us faster broadband conx with redundancy and a few more years for cloud based computing to get a decent foot in the door and I may be interested. Until then, not going to happen here.

[PiqueABoo]

There’s always DSL lines and Cisco of course, but not many schools want to spend that sort of money...

 

Schools, or at least the ones I encounter are apparently used to it - all have symmetric around here and increasing upstream flow for lots of different credible reasons is why. Many of those can't or won't be replaced by cloud-fluff any time soon, which is explicitly acknowledged in this comment of yours i.e. there is a "lot more" data at the school:

 

There’s also the security aspect of using VPNs to consider – if that bridge is breeched, then intruders have access to a lot more of your data

 

If I breach a cloud service I might get access to hundreds of thousands of folk's data. Only time will tell but with hindsight we may decide that some school with their random VPN occasionally being totalled would have been a better overall deal - think of it as a minor variation on the old but still perfectly good "Windows monoculture is bad" argument.

 

I suspect we might see something of a cloud backlash, in fact there are creeping signs of one already as a direct consequence of the script kiddy self-promotion/journo-publicity over the last couple of years. People here and there are talking about getting their important stuff off the net, back behind the air-gaps.

 

ports 80 and 443 used by web servers will always get high priority from any broadband traffic shaping.

 

Will be interesting to see what, if anything, falls out of the current CoE push on net-neutrality in that respect e.g. "Users should have the greatest possible access to Internet-based content, applications and services of their choice...".

Edited June 25, 2011 by PiqueABoo
tpyo

[BromcomPublicRelations]

In the 1990s when Windows replaced DOS, some software vendors tweaked their DOS applications to sell them under Windows without true the benefits of the Windows GUI, multi-tasking, multi-threading and host of other underlying features etc..

 

Now 20 years on, in 2010s Windows suppliers will be trying to repeat the history and sell their Windows based applications as “Cloud” solutions where terminal services is under the browser without the true benefit of the Cloud and at the expense of performance, higher server costs, flexibility, dynamism, larger carbon foot-print and underlying host of features etc...[/

 

 

Cloud Computing: Browser apps vs Windows apps:

 

Further to the above post, here is a new argument to drive the point home about some software vendors trying to sell Windows apps suitable for Cloud deployment:

 

The world’s largest software vendor for Windows applications is Microsoft. In recent days there has been a great deal of news and assumptions on marketing strategies about Microsoft’s push for the Cloud and seeing Office 365 as leading this.

 

Clearly Microsoft with billions of dollars of investment in Windows Office 2010, did not try to offer users all over the world the ability to run Windows Office 2010 application within a browser and call it Cloud deployment! Nor did they try to offer hosted web extensions to Office 2010 and call it Cloud deployment. Microsoft developed Office 365 ground up specifically as a browser based product suitable and optimised for Cloud deployment Office 365.

 

This example of Microsoft and Office 365 is clear evidence of what true Cloud apps vendors are doing and should be doing! The begging questions is - why didn’t Microsoft host Windows Office 2010 Windows application in their server farms and make it available to subscribers within a browser and call it Cloud service?

 

Cloud Computing and Cloud Deployment is still terms that are often misunderstood. Microsoft Office 365 and similar applications will educate the market to a level where soon vendors with Windows apps will be embarrassed to claim that running centrally hosted Windows apps with user access through a browser as meeting the ‘Cloud Deployment’ definition. Yes, they can be hosted but they are not Cloud and Microsoft knows it and so should all software vendors.

Edited June 29, 2011 by BromcomPublicRelations
typo

[pcstru]

Cloud Computing and Cloud Deployment is still terms that are often misunderstood.

 

Err ... Right. Although your post makes it clearer perhaps on what YOU think cloud deployment is not, it's not exactly shining out as to what it actually IS. And that's just what YOU think. 'cloud' is just another marketing term in an industry that grazes, masticates, digests and then milks such terms with a frequency that makes promoters of 'personal care' products feel positively lazy. Take 10 different sales bods from 10 different companies and they will sell you 10 different interpretations of 'cloud', 17 if you ask them back after lunch to remind you what they said in the morning session (23.4 if it was a really good lunch)- and all no doubt telling you that the other 9 are wide of the mark. Perhaps one day the 'cloud' will rise to match the rosy glow a majestic sunrise throws on early morning clarity and we will all bask in the warming glow. Right now it is so low lying that the experience is just thick fog.

 

I was considering Sony's recent experiences and the organised attacks against various providers from anarchic 'hacker' groups and it put into focus another risk of 'the cloud'. If you take some "platform as a service", outsourcing the iron to the likes of Rackspace who host for many other organisations, you are then opting into a risk pool where if one of those organisations is subject to a denial of service attack, the DOS may actually hit any number of unintended targets - including your little process that's running on the same iron as some scaled slice of the target. And we haven't yet really explored the vulnerability of hypervisors to well targeted attacks via the hosted processes. Perhaps you think your 'server' guests are well managed and hardened against attack but the guests next door are being run by ... well, who knows. Perhaps someone with too much time on their hands is already working out how to exploit buffer overruns in the installed VMTools to wreak their dastardly plans on the innocent hypervisor management layer and from that guest, own the host and all the other guests. Perhaps. Maybe.

 

The point is risk. Risk, probably one of the most poorly understood management disciplines, is measured not just by probability but by the consequences of that. Ask the Japanese who seemed to think 1 in 1000 year events were so improbable that the consequences didn't even need to be factored. My view is that there is a huge risk when we all run to 'the cloud' thinking it a secure environment (and it appears to be relatively secure but is relatively untested in that) but we are all pooling into the same pond of consequence.

 

Several years ago I found a little bug in MySQL. Unfortunately I didn't find it in the development system (I know, big minus points to me, but there you go) I found it in production. The production system was running on a fairly major ISP's servers, where the MySQL database was hosting databases for many hundreds of different websites. The bug bought the DBMS and the underlying servers to their knees. Many websites down - Film at 11. Now we have the likes of rackspace, tying thousands of servers into a single unified management infrastructure. Able like magic to scale the cores they can throw at a job from a slice of a single processor to thousands of processors across warehouses full of iron even on different continents. If nothing else, that's a huge big juicy target. Sometime, someday, someone will take a shot and score a hit. All your base will belong, all.

Edited June 29, 2011 by pcstru

[BromcomPublicRelations]

Err ... Right. Although your post makes it clearer perhaps on what YOU think cloud deployment is not, it's not exactly shining out as to what it actually IS.

 

PCSTRU,

 

You are spot on. We are specifically saying that just as running DOS application under Windows did not make legacy application shiny Windows applications with full benfit of Windows features, deploying Windows application under hosted environment will not make them shiny Cloud applications with full benfit of Cloud features and efficiency.

 

That is said – we need to cast our minds to a typical Local Authority that is minded to deploy MIS application for say 100 primary and 20 secondary schools on Cloud. We invite all to forward estimate of resources that will be needed in terms of processor and RAM etc.. for two scenarios

 

A. If MIS applications is in browser and optimised for Cloud

B. If MIS application is in Windows as a thick client

 

If not exact number of processors and RAM, please forward an estimate % less resources needed between the two scenarios in (A) vs (B). Also think of who is paying for the extra huge cost differnce, generated heat and carbon foot-print?

 

Answers on a postcard please.

 

NB: This is why Microsoft would have never dreamt of putting Windows Office 2010 as a Cloud offering and has written ground up a browser version for Office 365 as a Cloud offering.

Edited July 1, 2011 by BromcomPublicRelations
typo

[BromcomPublicRelations]

Err ... Right. Although your post makes it clearer perhaps on what YOU think cloud deployment is not, it's not exactly shining out as to what it actually IS.

 

A true cloud deployment is one that allows "all" the benefits of cloud computing come through. One of the key and overwhelming benefits is the efficiency gained through browser applications. In Windows based applications the resources needed in infrastructure in terms of server resources consumed to provide performance that is required by the schools is no less than in conventional (non-cloud computing) deployment. Whereas because the 'native' technology for 'centrally hosted' deployment/cloud computing is browser, there huge efficiency that flows when browser application is used in terms of the initial infrastructure cost and resources consumed in running.

 

Resource usage Web verses Windows in a hosted environment:

Web applications would typically have a single logical instance of the software application running on the server, sharing that executing code with all users, including across schools in this case. As users log in and out, ‘sessions’ are created to hold information related to that user, and those data are used by the shared executing application and discarded when no longer required.

 

The hosted Windows applications case is very different. No longer are users sharing the executing application with other users. Instead the host needs to create a virtual environment for each user in order to execute a separate application instance for just that user and hold their data and other operating overheads. Often a virtual desktop is created and the user in effect has a virtual machine at their disposal. So, on top of the sessional data for the user required for the web application, the hosted windows application has far higher demands on the server resources and needs to contend with the overheads of running multiple copies of the windows application in a virtualised environment, which itself introduces a further abstraction from the server hardware at the data centre.

 

Extra hardware can of course be “thrown” at the challenge of hosting Windows applications, but performance is not necessarily predictable when the application is put into an environment is often will be alien to that considered during design, development and testing for such applications.

 

Invariable Windows based applications deployed in hosted/cloud environment ends up with highly bloated infrastucture and naturally very high cost in comparision to browser based application!

[pcstru]

I'm struggling to see "web based" as "cloud" or as in your previous post parse "Cloud applications with full benefit of Cloud features and efficiency". It all just sounds like marketing bull - from someone who uses the word "cloud" because they are a bit afraid of talking about Christmas to turkeys and using the word they actually mean, "outsourcing".

 

I'll gamble a definition of "cloud" in a bid to advance the discussion a little; at least then people can chip in and say why I'm wrong. This is as simple as I can put it but captures the key essence :

 

Cloud Computing : Decoupling the platform from the constraints of the underlying hardware.

 

A web platform can be run on windows hardware and in a schools environment, the clients are likely to be sitting on windows PC's. Then benefits you cite for the cloud are not cloud benefits, they may not even be real benefits - only if you make major changes to infrastructure will you realise them and that has nothing to do with the application as such.

 

"Cloud applications with full benefit of Cloud features and efficiency" - applications can take advantage of the platform - there is no cloud API that applications can tap into. It's the platform that allows true cloud efficiency. That efficiency is in the way that the platform can scale from ticking over on a few cycles on a single core, to supporting hundreds of thousands of users and seamlessly spreading the load over thousands of cores on machines perhaps located in different continents. That should all be transparent to the application - and is hardly something that should be bothering a schools MIS system which might be asked to support at most a few hundred simultaneous users (well, OK, SIMS Discover might benefit from a few thousand cores to draw a pie chart now an again), something that can readily be achieved on relatively inexpensive hardware with a virtualised platform (if you are after efficiency of the iron) that has no capability to scale over more than a few cores. "web based" is not cloud - it's just ... "web based".

Edited July 6, 2011 by vikpaw
language

[GrumbleDook]

Another thing that gets me with the term 'cloud' is that I have had a sales drone ramming down my throat that cloud applications cannot be run anywhere other than out on the internet. The idea of running a services from within an RBC data centre or running them at another school or even running them within the school infrastructure itself.

 

"Cloud" is just another bunch of marketing speak ... the same web 2.0 have been.

 

It is also being done as an attack on locally or centrally hosted applications which might happily do the job that is needed ... and doing it well ... but because it is not in the "Cloud" ... it is wrong.

[BromcomPublicRelations]

I'm struggling to see "web based" as "cloud"... - from someone who uses the word "cloud" because they are a bit afraid of... actually mean, "outsourcing".

 

"web based" is not cloud - it's just ... "web based".

 

pcstru,

 

We appear to have cross lines.

 

1. We aren't claiming being "web based" means "cloud". If you have "cloud" infrastucture (we made no attempt to define this hereto - except an extact quote below), then what we are saying is that you need to have a browser based MIS rather than Windows based MIS to get the efficiency. We have provided reasons in reasonale details to explain why in our previous post.

 

2. "cloud" of course generally means "centrally hosted" service which is outsourced. That should be seen by some as part of the benefit - specially for primary schools with no ICT support staff or budget!

 

As to your definition " Cloud Computing : Decoupling the platform from the constraints of the underlying hardware." - I have no problem with this.

 

A definition: "True Cloud"

This forum may may benefit from an interesting definition below which is an extract from a recent tender requirement. It is intersting to read quote "true cloud" and the fact that someone is "putting their money where their mouth is" to this definition!

 

A true cloud computing service:

 

Cloud computing services can be deployed in a number of different ways, with varying internal or external ownership and technical architectures. To be considered a true cloud computing service, certain core principles need to be adhered to:

 

— technical abstraction: Users of services have limited or no visibility of the technology (both hardware and, where appropriate, software) used to provide the service,

 

— pay per use: Costs are incurred as revenue rather than capital expense, based on a utility “metering” model,

 

— rapid elasticity: Service volumes are elastic, being readily scalable both up and down in volume to suit ever changing business demand,

 

— ubiquitous network access: ICT services are made available over the network using standard mechanisms which support a wide range of client devices (e.g., desktop, laptop, PDA, phone).

 

The main categories of cloud computing service:

Cloud computing services are varied and maturing and include services that span all layers of the traditional hardware and software stack. The three main categories of cloud computing service are defined as:

 

— infrastructure as a Service (IaaS): The provision of lower level basic server and storage capabilities,

 

— platform as a Service (PaaS): The provision of a computing platform combing hardware and software that can be used, for example, to support application development or testing activities,

 

— software as a Service (SaaS): The provision of a complete application offered as a cloud computing service over the network.

Edited July 5, 2011 by BromcomPublicRelations
typo

[pcstru]

pcstru,

 

We appear to have cross lines.

 

1. We aren't claiming being "web based" means "cloud". If you have "cloud" infrastucture (we made no attempt to define this hereto - except an extact quote below), then what we are saying is that you need to have a browser based MIS rather than Windows based MIS to get the efficiency. We have provided reasons in reasonale details to explain why in our previous post.

 

I don't think we have crossed lines. You are wrong in equating 'efficiency' with browser based; either that or you have failed to define efficiency sufficiently. What you get with browser based is (or should be) independence from the client operating system (note, not the client platform). In these days of AJAX and rich web applications supported by huge munging chunks of javascript, it would be quite possible to put together an application that had a smaller client footprint than one using a browser (perhaps using ... Java). And there is no reason why that application can't run on multiple platforms.

 

I like the specification for cloud from your client though, it's good criteria on which to judge true 'decoupling'; what they are saying is just give us the service, charge us on use and frankly we don't much care how YOU deliver it.

 

In the end I guess I just get a little jaded by the endless sausage factory of 'technical' terms which are spewed out by or latched onto by marketing departments without much appreciation of the technical reality. We end up with bucket loads of largely meaningless word salad when what they are actually trying to sell is just yesterdays technology dressed up in the emperors finest. I feel like the little boy exclaiming in the crowd - "he's stark streaking naked people!". I have to take succour in the small irony that "cloud" must at least have cast a shadow on the "blue sky" of the "brainstorming" sessions and imagine the marketeers huddling under it to at least get into the cloud precipitated "idea shower". Next - thinking outside of the "umbrella" - "embracing the light drizzle beyond the cloud". Oh, how I do so look forward to the future.

Edited July 5, 2011 by pcstru

[matt40k]

— technical abstraction: Users of services have limited or no visibility of the technology (both hardware and, where appropriate, software) used to provide the service,

 

— pay per use: Costs are incurred as revenue rather than capital expense, based on a utility “metering” model,

 

— rapid elasticity: Service volumes are elastic, being readily scalable both up and down in volume to suit ever changing business demand,

 

Well put.

 

— ubiquitous network access: ICT services are made available over the network using standard mechanisms which support a wide range of client devices (e.g., desktop, laptop, PDA, phone).

 

I don't really agree, it sounds like Web rather then cloud, after all I could create app that only runs on Desktops, but it would be using Amazon\Azure\RackSpace therefore it meets the other requirements.

[synaesthesia]

"Cloud" is just another bunch of marketing speak ... the same web 2.0 have been.

 

It is also being done as an attack on locally or centrally hosted applications which might happily do the job that is needed ... and doing it well ... but because it is not in the "Cloud" ... it is wrong.

 

Here here. As I've said before, it irks me that people use it as a new term. SIMS has been in "the cloud" longer than "the cloud" has actually existed.

Most of our services are provided over the cloud. By us. VLEs, email, MIS - oh hang on; that's everything. Because "The Cloud" could apply to *every single part of the internet and WAN connectectivity in general*

 

The sooner people stop calling it The Cloud the better. It doesn't matter how some people will dress it up, the definition of "The Cloud" is simple. Something not local. So far it's entire use is to con people out of something they can probably already get just because it has a fancy name attached to it.

 

Remember that next time you outsource making your pizza to the local pizzeria and have it delivered via The Cloud

 

 

OOoh, edit - they should call it "Pea Soup" or "The Fog" - because it's synonymous with having your eyes covered instead of opened. "You can't see it or touch the server" etc etc - that's right, and also no control over when things go pear shaped or the ability to smack it with a handy piece of 2x4

Edited July 6, 2011 by synaesthesia

[BromcomPublicRelations]

Most of our services are provided over the cloud. By us. VLEs, email, MIS -

 

The sooner people stop calling it The Cloud the better. It doesn't matter how some people will dress it up, the definition of "The Cloud" is simple. Something not local.

 

"Something not local" & efficiency of MIS running:

 

As stated previously we will not dwell on definition of Cloud. This thread has the title of cloud - hence why discussions appear to be going about teh definition.

 

Where as we have been advocating discussion on the efficiency.

 

On the efficiency, we challenge anyone questioning the significant difference in scale of efficiency running "Windows application / thick client based" MIS vs browser based MIS over centrally hosted/"something not local"/Cloud infrastructure/service. What we mean by "Windows application / thick client based" is client-server software that presently run by probably 90% of UK schools.

 

The reason for this is simply the fact that "Windows application / thick client based" MIS software resource demands moved from client PC/workstation to centrally hosted infrastructure in verbatim manner eg having private copy of the application run for each user. Whereas browser application are developed and optimised to share instances of application code across all users (100s/1000s/millions !!!).

 

More detailed explanation is given in earlier posting and for convenience and completeness copied here:

 

Resource usage Web verses Windows in a hosted environment:

Web applications would typically have a single logical instance of the software application running on the server, sharing that executing code with all users, including across schools in this case. As users log in and out, ‘sessions’ are created to hold information related to that user, and those data are used by the shared executing application and discarded when no longer required.

 

The hosted Windows applications case is very different. No longer are users sharing the executing application with other users. Instead the host needs to create a virtual environment for each user in order to execute a separate application instance for just that user and hold their data and other operating overheads. Often a virtual desktop is created and the user in effect has a virtual machine at their disposal. So, on top of the sessional data for the user required for the web application, the hosted windows application has far higher demands on the server resources and needs to contend with the overheads of running multiple copies of the windows application in a virtualised environment, which itself introduces a further abstraction from the server hardware at the data centre.

 

Extra hardware can of course be “thrown” at the challenge of hosting Windows applications, but performance is not necessarily predictable when the application is put into an environment is often will be alien to that considered during design, development and testing for such applications.

 

Invariable Windows based applications deployed in hosted/cloud environment ends up with highly bloated infrastructure and naturally very high cost in comparison to browser based application!

 

 

synaesthesia,

How about sharing with this forum, the number of primary and secondary schools in your LA you run MIS "not local" and what hardware resources you had to dedicate to do so?

 

We will be happy then to estimate hardware saving you can have if the MIS application was in browser.

 

I am sure this forum will benefit from some factual data.

[synaesthesia]

@BromcomPublicRelations: Think you've gotten your wires crossed somewhere, I have no links whatsoever to my LA, my views are my own entirely etc. I used to support a number of primary schools on behalf of a larger support provider and had many fingers in many pies when I did so, hence the experience especially with SIMS. I was not working with any MIS other than SIMS, and when SIMS wasn't run locally it was only in very small primaries where hosted solutions were a better financial bet.

 

I am more than aware of the savings that can be made, but money is far from the point; what good is saving £5k when you're not getting what you need. I'm not saying that's true of your product or anyone elses, just saying there's a lot more to it and I start to get the feeling you're seeking opportunity to try and sell your product a little too much

 

Personally, I feel that hosted solutions are *not* yet a worthwhile move for MIS unless you're sitting on top of your carrier, most certainly for larger establishments. The risk is still considerable when you have several departments entirely reliant on the reliability of your MIS. Too many points of failure for too little benefit.

[BromcomPublicRelations]

I am more than aware of the savings that can be made, but money is far from the point; what good is saving £5k when you're not getting what you need.

 

The arguments we put across about “Resource usage Web verses Windows in a hosted environment“ is for an LA who already decided to go ‘centrally hosted’ service and why they should consider ‘web based MIS’ rather than ‘thick client Windows’ for this. You may be surprised that a good number of LAs already have gone for ‘centrally hosted’ service for their schools for some years! Certainly we believe it makes sense for primary schools – so long as broadband is up to it! We appear to agree that "there are saving to me made" in doing so. However there is significant more saving in infrastructure that can be made if browser based MIS is used. Many more LAs are expected to consider “centrally hosted/cloud” option in their annual reviews how to accommodate the spending cuts.

[matt40k]

Are you saying a MIS written from the ground-up as a MIS application, or an existing well designed MIS with a web front end?

 

I'd be interested to hear how you would deal with say, end of year reports for the entire school, we must be talking a few hundred MB. I admit the cloud resources would make generating them alot easier, but printing might be interesting, still I suppose you could do e-reports. Also aren't certain features going to be more difficult on a web browser?

[GrumbleDook]

The arguments we put across about “Resource usage Web verses Windows in a hosted environment“ is for an LA who already decided to go ‘centrally hosted’ service and why they should consider ‘web based MIS’ rather than ‘thick client Windows’ for this. You may be surprised that a good number of LAs already have gone for ‘centrally hosted’ service for their schools for some years! Certainly we believe it makes sense for primary schools – so long as broadband is up to it! We appear to agree that "there are saving to me made" in doing so. However there is significant more saving in infrastructure that can be made if browser based MIS is used. Many more LAs are expected to consider “centrally hosted/cloud” option in their annual reviews how to accommodate the spending cuts.

 

But Why would an LA take on a centrally hosted option in the present situation? Unless they can get long-term guarantees from schools and academies that there will be buy-in then it creates too much of a risk. Even if a school signs up now ... when they become and academy they might decide that it no longer applies and they go elsewhere ... leaving the LA with a cost to manage ... and that either gets absorbed by the LA or spread over schools.

[matt40k]

Have a word with Cambridgeshire LA, they've not long ago replaced there kit.

[GrumbleDook]

Have a word with Cambridgeshire LA, they've not long ago replaced there kit.

 

All of that was sorted / arranged prior to DoE and Mr G deciding to move the goal posts, IIRC. They have also just signed a deal with Virgin on connectivity too (CPSN) which should make quite a difference to school connections.

[BromcomPublicRelations]

Are you saying a MIS written from the ground-up as a MIS application, or an existing well designed MIS with a web front end?

 

I'd be interested to hear how you would deal with say, end of year reports for the entire school, we must be talking a few hundred MB. I admit the cloud resources would make generating them a lot easier, but printing might be interesting, still I suppose you could do e-reports. Also aren't certain features going to be more difficult on a web browser?

 

matt40k,

For the efficiency, MIS software must be ground up written as web application so that a single logical instance of the software application runs on the server, sharing that executing code with all users, including across schools in this case. As users log in and out, ‘sessions’ are created to hold information related to that user, and those data are used by the shared executing application and discarded when no longer required.

 

Web front end of course will behave same as above however when Windows part is fired up the benefit of web front is lost.

 

Year reports for the entire school:

Such extensive reports have been produced by hundreds of school using all web based Bromcom e-Registration system since early 2002. Printing, resource management for local or system printers are all highly configurable.

 

… aren't certain features going to be more difficult on a web browser?

We are not sure which features you mean. Yes, we had some serious challenge but mastered over the past 10 years how to incorporate wide range of features - such as ‘right click’ functions, all sort of cursor auto movements, intelligent selectors, remebering last used names etc.. Every report in the system can be scheduled to be generated automatically and printed on any printer in the system or e-mailed to a chosen receipent at a scheduled given day of the week, month, year at chosen time and frequency. All these also are built in Bromcom MIS. We now receive comments from schools saying ‘your MIS is so intituive - easy to use as Facebook” – assuming it is a complement. Please drop us PM if you wish to talk to schools using all-browser MIS.

 

Please state any particular feature you like to check - happy to explain.

[BromcomPublicRelations]

But Why would an LA take on a centrally hosted option in the present situation? Unless they can get long-term guarantees from schools and academies that there will be buy-in then it creates too much of a risk. Even if a school signs up now ... when they become and academy they might decide that it no longer applies and they go elsewhere ... leaving the LA with a cost to manage ... and that either gets absorbed by the LA or spread over schools.

 

GrumbleDook,

 

An interesting and a good point.

 

At present time most primary schools appear to be sticking with LAs. With DfE-Buying Solution Framework OJEU Tender for MIS and Learning Services which is now well underway, we may see a different market structure being emerged for supply and services for MIS & Learning Services. Under new Framework contracts LAs may want to negotiate terms with software suppliers so to allow small % of drop out of primaries without a penalty. With "software as a service" (SaaS) and utility model service discussed under centrally hosted/cloud on this thread, a school should be able to switch support services/delivery/billing from the source just as billing agent for domestic gas and electricity. Hence a school may switch from LA to other without disrupting the service if stays with the same software. Changing to a different software is a different matter of course.

 

Meanwhile LAs/schools will check how they will cope with spending cuts and evaporated Harnessing Technology Grant and see how much centrally hosted MIS may help to reduced deployment/support staff and lost budget to keep MIS servers up to date.

[GrumbleDook]

@BromcomPublicRelations

As you know, few LAs (if any) specifically used to pay for ongoing costs for hosted services for MIS. They may have done for Learning Platforms (as part of capital projects to get things running) or to enable MIS - LP integration ... but there have been enough FoI requests going round to show that pretty much all of those contesting the majority providers position have that information now, as well as the papers, conslutants and many more.

 

Changing from an LA hosted central service to a cloud based SaaS under the new framework will be an interesting choice for many. The analogy of viewing it as a domestic service is not quite right though ... depending on how the centrally hosted services is provided, accessed and supported it is likely to differ from SaaS offerings due to different UI, functionality, and support ... even if it is from the same generic supplier. The supplier has to persuade the LA to accept the change of service to the new style service, to take on board the possible costs for this (ranging from different hardware requirements, support requirement, software and licence requirement and training requirements) and to migrate to a new service which would allow for easier migration to a SaaS service which would be available under the new framework (when it arrives).

 

Should a school leave the LA service at this point the LA could have reasonable chance to refuse to pay any penalty to the supplier. That is a position being looked at under the new framework, I believe, as transition is going to be a stumbling block for a free and open market. Transition and Migration have costs ... even moving from a like for like service.

 

I will be interested to see the commercials on the new framework to see what discounts will operate for LA or large scale purchase. Whilst SaaS is viewed by some as a reason why there should not be a discount based on scaling ... this is fine for the hardware / computing aspect ... but fails to deal with support / training. Still ... that is up to the DoE to look at and I am sure that it has been considered properly.

[synaesthesia]

"MIS - easy to use as Facebook ... a complement?"

 

That caught me off guard a little bit.

 

Sorry, but hell no. If an MIS is "as easy to use as Facebook" then it's doing something wrong. If that was the case, just about anyone can rock up to it and get their hands dirty, breaking god knows what in the process. MIS administrators should be specifically trained people to do a particular job; the thought of randoms being able to get their hands onto it scares me - I think that's also a little on the dangerous side. After all, everyone would like to own a car that you can fix yourself but there's damned good reasons to trust qualified people to do it for you. That would seem to be another argument though, not for this thread.