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

[Mike_Bostock]

There are quite a few comments on Cloud Computing dotted about these forums but I was wondering what the current thinking from technical staff is on Management Information Systems that are hosted off site.

 

We will soon see Progresso from Serco as the latest example of a Cloud MIS, so I expect that they will have worked on solutions to many of the obvious constraints like security, service availability and data transmission speeds.

 

The key plus points seem to be ease of maintenance and availability from anywhere there is a browser. So far, in our work we have taken the view that a school-installed system is more secure, if it is written well it should be low maintenance, and if folk wish to use it from home they can use VPN - and get access to all their other stuff on their PC at the same time. Also, if it is school-based it might be easier to transfer the data to and from other data tools.

 

What is the current thinking about these issues? Is there an urgency for applications to move to The Cloud? By that I mean, will schools soon be moving to choosing solely Cloud based solutions to data management rather than school-installed applications?

[matt40k]

Personally it's a personal choice. It's based on the head teacher, the location and size of school.

 

If a head wants a "cloud" MIS, he's more than likely to get it. If they're in an area where the fastest broadband is 1MB, it's unlikey compared with a school who's got a 100MB. Finally size, it doesn't make sense for a large school with hundreds of users to use the cloud. Having said that it I'm assuming cloud is remote, but Azure allows you to sync local and remote resources together so I guess that would be cloud. So I guess that leaves network and head teacher, and with faster more reliable broadband that leaves just the head and well with decreasing funds I guess cloud might be the answer. I just don't see it happening widespread for a number of years.

[dhicks]

I expect that they will have worked on solutions to many of the obvious constraints like security, service availability and data transmission speeds.

 

But the times when you might most need service availability - e.g. when there's a fire and you need to take a register - would coinide with the times when there's most likely to be a network outage.

[matt40k]

I was playing around with building a web based register program for SIMS, started out as just a web version of the SIMS register, but someone pointed out it would be good it it was on a remote web server so you could access it even if the onsite SIMS server was on fire... from your iphone

[elsiegee40]

As a small school with limited onsite resources and technical time we are seriously considering cloud for MIS and other applications. I know it will sound astonishing to most schools that we don't have a formal MIS at present; we're in the process of examining our options now... first salesmen across the threshold next week.

 

We have neither budget nor staff capacity to provide the level of software, servers and access that we would like in-house. My role becomes more administrative looking after the local network and we can have the software with all the bells and whistles on hardware that is somebody else's problem and both get upgraded and updated without worrying our tiny budget too much.

 

So much data in schools is already 'cloud' already, e.g. live@ edu/office 365, with sensitive data entrusted to the servers of others. Personally I see it as becoming the norm, not just in schools, for all except those businesses with the cash to be able to provide a 24/7 service that can be accessed from anywhere in the world.

[ntm1275]

Our LA went down the cloud MIS three years ago, and it is the best thing we could have done

 

Everything is located centrally, meaning schools do not have to manage servers, or software and updates

 

It is also considerably cheaper than SIMS

[PhilNeal]

We won our last two LA tenders by providing SIMS as a cloud solution. The one before we won because SIMS could be deployed locally. It all depends on your connections.

[dhicks]

it was on a remote web server so you could access it even if the onsite SIMS server was on fire... from your iphone

 

Good point. Also, if your MIS does timetabling, it might be handy to have a distributed, high-speed compute cloud available to reduce the runtime for timetabling algorythms.

Edited May 13, 2011 by dhicks

[JPS]

Good point. Also, if your MIS does timetabling, it might be handy to have a distributed, high-speed compute cloud available to reduce the runtime for timetabing algorythms.

 

I think the Serco offering (Progresso) will still have a locally installed solution (Scheduler) for timetabling with.

[Sivadam]

Methinks that we will have to get a darn sight better Broadband connection to work solely in the clouds. YHGfL installed their latest super-dooper offering over Easter. At times my old 256k modem was much faster! And I am not kidding!

All my Timetable (NT6) files are saved to DropBox but it is a good job they sync to my PC at 7am every morning when I log on, else they would never be up-to-date! The connection is appalling during the 'normal' working day. If we were all running around in the clouds then we would be able to get 'nowt done!

Edited May 13, 2011 by Sivadam

[pcstru]

What is the current thinking about these issues? Is there an urgency for applications to move to The Cloud? By that I mean, will schools soon be moving to choosing solely Cloud based solutions to data management rather than school-installed applications?

 

I think it will depend on the School. As a large school we have dedicated MIS support on site that deals with hosting and usage issues. On the usage, the local knowledge (the way we use certain features or interpret certain data) is highly valuable; but that doesn't account for 100% of the role, so the fact that upgrades etc are done makes the role reasonably efficient. The local knowledge is extremely well placed to deal with competing demands and make decisions based on priority that is appropriate for the situation - it's not just a bunch of calls in a queue and it's not competing with other schools. As a large school we have a lot of users - several hundred staff may be using the system simultaneously and the local infrastructure is well placed to cope with that. Shunting all that over the broadband pipe introduces additional latency and increases the demand on our firewall and filters. What we might gain in removing some of the self service element (which let's face it are mostly staff costs) we may need to pay out as investment in our broadband infrastructure to cope with a) additional demands and b) change in rationale for broadband. Finally there is ownership and access to the data. There are already examples of cloud providers losing critical data. That runs from people using flickr to store their precious photo's to organisations who have paid for off site backup that wasn't (recent example was of a film production company losing a load of footage they had decided to store in the 'cloud'). MIS data for a school is critical; so if it is going to be hosted elsewhere, then we need to be expert at due diligence when it comes to engaging those partners. That's an expertise I don't think is mature yet in any sector, and education is no different.

 

But it depends on the school, I can see that small schools might benefit. Some of our local primaries don't really seem to have good backup in place, they might be better of (or certainly no worse off!). They don't have dedicated onsite support anyway so have little to lose there and concentrating on good broadband provision might actually start top open up other useful services for them.

 

Even for us perhaps things will change. However I'm reminded of the mainframe days when the PC came along. The PC quickly displaced the mainframe for many reasons but one was particularly prominent. The PC put the power in the hands of the end users rather than the central and often unresponsive 'computing department'. Suddenly instead of having to wait months for a report to be written and run as a job, people could do it themselves (or at least the geek in the corner could). To some extent I feel 'cloud' services involve giving up a lot of control and that may be something people will realise that they miss, only after they have made the move.

[ntm1275]

From experience, there can be huge differences in speeds depending on the browser that is used for the cloud MIS

 

All version of Internet Explorer are pitifully slow, including IE9 which Microsoft claim is their fastest ever browser

 

Firefox is faster, but the champion at the moment is Chrome

 

I have found that Chrome can display the MIS at up to 70% faster than IE can

 

Even on a 0.5Mbit connection, Chrome makes the cloud MIS feel like it's installed locally, where as IE needs at least a 4Mbit connection to be comparable

 

Cloud MIS's can be fantastic for Primary schools that do not have dedicated IT staff, as all server hardware, software and backups are managed centrally - also saving on the cost of separate hardware

 

From dealing with Secondary schools, cloud MIS's can take away the burden of managing the technicality of the hardware and software, so that the IT staff can get on with managing other school IT systems, which there are plenty

 

Yes there is the question of the Internet connection going down, but our Secondary schools have backup connections to cover this possibility, which 'touch wood', they haven't had to use

Edited May 13, 2011 by ntm1275

[glennda]

I think the idea of a web-apps type thing is good. although i wouldn't want to have it hosted externally. 1. i don't trust the internet connection enough at the moment - once its gone everything is gone (being part of an LEA wan with one connection only) 2. We are still stuck with a 10mb connection for the next 2 to 2 1/2 years without room for improvement to 100mb. 3. the data protection part is to much of a risk for me if data is lost on my server then is my fault (and i have full control over keeping it safe) if the data is lost/comprimised on googles servers then its still my fault... the likes of how at risk things can be is highlighted by 100 million peoples data being lost by Sony.

 

I would rather have a group of clustered webservers onsite which will host Office apps (although i would like an open source/freeware solution rather then MS).

 

And for MIS all these risks are just incresed beyond an acceptable risk for me

[PiqueABoo]

I'm niggled by the reverse of a problem I got to fret about in the first 1/2 of the 90's: Adding Internet to WAN links built for a critical business app. Here & now it's adding a critical business app to WAN links used for Internet access. Same as back then I think I want "prioritisation" for the critical traffic, otherwise a lot of kids simultaneously accessing "heavy" net content can seriously harm/randomise an MIS user's experience, especially in a lower bandwidth site e.g. Primary. Are folk leaving all this to chance or attempting to do something about it?

 

As discussed in one of the older threads I'm not sure I'd be entirely happy with cloud-MIS without some kind of light-weight fallback service that could cache critical stuff locally e.g. today's attendance, should the path to the cloud be down (which is what the 90's business app did). Trouble is the chances of getting lightweight anything from anyone these days seem very poor.

 

Primary schools saving on the cost of separate hardware

 

They don't need that, one server and a dash of not that expensive extra RAM, one VM for jack-of-nearly-all-trades school DC, another for bloated MIS. It's a bit of a no-brainer, but IME quite rare. Increased S/W complexity (your tech needs to grok Hyper-V,/ESXi/whatever) offsets the h/w reduction a bit though.

Edited May 13, 2011 by PiqueABoo
seperate h/w

[john]

I think the Serco offering (Progresso) will still have a locally installed solution (Scheduler) for timetabling with.

 

Correct, to start with you will still use Scheduler for Timetabling, however you will be able to use Keith Johnson's timetabler as well and integrate that into Progresso, so the choice is yours. But its still based on the extraction and integration just like Facility is now

[penfold_99]

@john You will be able to use SIF to transfer data between Progesso and Keith Johnson's Timetabler.

[Mike_Bostock]

Thanks for all the responses so far to this question about whether Cloud Computing will become the required format for school Information Management Systems.

There seems to be an equal number of pros and cons to having a school's data managed offsite and delivered to a school through a browser. It certainly feels like there isn't yet a definitive answer to the question, so I look forward to seeing the discussion continue.

 

Probably one of the reasons for the lack of a yes/no answer is that I didn't start by defining my criteria and then asking for advice on a best-fit solution. Perhaps I should have said I wanted a solution to managing my information that was effective, low-maintenance, economical, accessible to all users, extensible, interoperable, adaptable to changes in legislation, etc. But even then I suspect that the answer would depend on individual school circumstances. For example, if a school has a technical team with good skills then there is perhaps more choice than if it doesn't.

 

One of the things that is haunting me about Cloud solutions is that it sounds very much like the days of mainframe computers and dumb terminals. Of course, most of us are too young to know much about these things, but I am told that there was a lack of user empowerment and an absolute dependence on the quality of traffic on the wire that delivered it to you. Of course, things don't need to be like that with modern systems but I would suggest that questions about these two aspects of the Cloud will keep needing to be asked.

 

Historically, when processing power became cheap, PCs took over from mainframes. The case for Cloud has its example in things like search engines. By using Google we are using their network of thousands of computers to do a processing task that we couldn't do with our PC on its own. I suggest that we need to find similar arguments when we specify Cloud as the solution to other types of problem.

 

One problem that I think we have here with our specification is that we assume that the current function of MIS will always remain the same. The very name 'Management' Information Systems does suggest the passive organisation of data for managers to look at. I would feel that, at a time when money is particularly tight, the question of Value for Money will come to the fore. i.e. Does the system do enough for what we are paying out? In the case of MIS that needs to include both the cost of the system itself and the cost of the people who maintain and manage it.

 

I personally don't think that 'managing information' is going to be demanding enough for schools in the future. The whole point of collecting data must lie in what a school does with it and the impact that is having on the effectiveness of the school. For example, we might wish to judge a registration system on whether it reduces truancy and helps allocate additional learning to those who missed out through illness. In the area that interests me - performance data - collecting pupils' scores and colouring them red or green may be where many schools are, but it isn't going to impress a visiting inspector unless a school can show what the school did with this information and whether their processes are improving standards.

 

A while ago we put together a checklist of sorts to try to suggest to schools how far they needed to go in moving from a 'functional' use of data, to a 'systemic' use - with a high impact on the work of a school. This checklist can be seen at this link. What is useful about something like this (and I don't claim it is the complete picture) is that it can help us be more specific about our requirements in order to answer the question about whether Cloud is a good approach in any particular case.

 

If I take an example, if a school wants all teachers to engage in the school's performance data and undertake research using it, then a distributed approach will become an important criteria. A school might then puzzle why it has a 'Data Manager' to manage the school's data when it doesn't delegate one person to do the breathing for everyone else. This may suggest a Cloud approach if teachers wish to make a lot of use of the MIS away from the school - or maybe a good networked solution coupled with VPN would have more advantages.

 

Will Cloud computing put limits on what you can do? Does browser technology have limitations over a standard PC application which become amplified when you are processing a lot of data? Which data technically needs to be processed locally and which remotely? What data will we need to cache? If we are working with data across several applications does Cloud become a barrier or not?

 

I could go on, but this reply would get even longer than it is.

So, for me the answer to Cloud lies in whether it meets the criteria I suggested earlier, avoids the twin problems of disempowerment or dependency on delivery speed, is not restricted in what it can do through using browser technology, and that the MIS solution chosen has an impact on the work of the school that can be weighed up against its cost.

Edited May 15, 2011 by Mike_Bostock

[plus4schools]

I think you have to differentiate between cloud based MIS and hosted solutions. Cloud MIS may run into Data Protection issues as the data could be spread around and stored outside the UK/Europe and be subject to different legislation. However, if you go down the hosted route, your data may be in the UK and governed by the UK Data Protection Act.

 

Bandwidth may be a problem if you try to run a client server MIS or one that was designed with client server architecture over the internet. But if you use something that's optimised for the web, then there's no problem.

 

Elsiegee40 - you're not alone in not having an MIS. Many independent schools of varying sizes rely on a combination of the school secretary, spreadsheets and database programs. Perhaps question your visiting salesmen about the location of the data to have the reassurance that it complies with the DP Act and that you can answer any parental questions.

[BromcomPublicRelations]

In 1990s when Windows replaced DOS, some software vendors tweaked their DOS applications to sell their DOS application under Windows without true benefits of 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 with terminal services under 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...

[CAM]

A "true" cloud system where you host the services on mutiple remote data farms: probably not. Too risky relying on an Internet connection and some third party to look after things.

 

A semi-cloud system run over an LEA Intranet: That would be better! One central MIS with automatic transfer of pupils from one school to another, centralised report system for parents with kids at multiple schools in the borough and support from the LEA IT team. And data isn't put into the hands of a strange third party. It is also not vulnerable to wider Internet faults (but beware of links to the LEA Intranet going down).

 

Unfortunately the idea of an LEA based cloud MIS will likely be killed off by Academies.

[GrumbleDook]

It is quite interesting to hear about the interest in the cloud at times ... and schools saying LAs need to move to it. Considering that many RBCs have data centres which host services and resources over a range of hardware and locations, the cloud has been used in education for some time. It is mainly the 'bolt a bit on' style of use (for scaling, etc) that is newer.

[glennda]

A "true" cloud system where you host the services on mutiple remote data farms: probably not. Too risky relying on an Internet connection and some third party to look after things.

 

A semi-cloud system run over an LEA Intranet: That would be better! One central MIS with automatic transfer of pupils from one school to another, centralised report system for parents with kids at multiple schools in the borough and support from the LEA IT team. And data isn't put into the hands of a strange third party. It is also not vulnerable to wider Internet faults (but beware of links to the LEA Intranet going down).

 

Unfortunately the idea of an LEA based cloud MIS will likely be killed off by Academies.

 

I wouldn't even trust our LEA with our Data - they are basically capita now!

[PiqueABoo]

the cloud has been used in education for some time.

 

Recall a conversation with someone around the turn of the century joking that the seriously buzzing Application Service Provider (ASP) thing would put me out of a job any day now - it's just another marketing term for the stupid & gullible, albeit one with a bit more "oomph" now.

 

The best ones are the schools that somehow think the "cloud" will currently wash away all their infrastructure and save them lots and lots of money.

[plus4schools]

Recall a conversation with someone around the turn of the century joking that the seriously buzzing Application Service Provider (ASP) thing would put me out of a job any day now - it's just another marketing term for the stupid & gullible, albeit one with a bit more "oomph" now.

 

The best ones are the schools that somehow think the "cloud" will currently wash away all their infrastructure and save them lots and lots of money.

 

Cloud computing has been around in the business world for a number of years in the form of programs like salesforce.com. Businesses require the same things as education does; security,value for money, ease of use, portability, instant access and so on. For individual businesses, these programs have made maintenance, upgrades etc of the program easier as it's taken care of by the supplier like Salesforce. Companies don't have to invest in their own technical expertise to make the program work.

 

The difference between these cloud programs and some of the old, current, educational ones is the way that they have been developed. If you have an MIS, or any educational program, that's based on old technology, as BromcomPublicRelations outlined above, then you are still going to have problems and require IT people to solve them.

 

If you use programs that especially written for the cloud or hosted approach then the requirement is reduced or eliminated. Some of our customers (and this is an example, not a sales pitch) don't have IT technicians, the program is just managed by the school secretary.

 

In some small, cash strapped schools, the infrastructure doesn't exist. Cloud applications like Google Education and the like are helping fulfil the ICT teaching requirements, but without the infrastructure.

 

That's not to say that IT expertise is not valuable in schools. What the best cloud and hosted services do is to free up IT specialists from getting bogged down in keeping the infrastructure working, giving them the time to do the much more skillful job of making sure the data is meaningful and contributing to the education of the students.

 

But it's horses for courses. There are some good programs that have been built for the internet but not everyone in education feels comfortable going down that route at the moment.

 

How about this for a barometer test: check how long until you next access anything on the internet whether it's a website, Google search, email, Facebook or whatever, using your computer at work, home or your smartphone. Think about how inconvenient it would be if you couldn't. I suspect the vast majority of us will conclude that we are already hooked on the internet and therefore cloud of one sort or another ...

[PiqueABoo]

The difference between these cloud programs and some of the old, current, educational ones is the way that they have been developed.

 

I started s/w engineering as a day job 20+ years ago and the way I see it, it's all "old" tech e.g. we had thoroughly web-ified a seriously non-trivial MIS system well before the turn of the century, and were very keen to do hosting but the customers were still struggling to grok the net, wanted the servers and data on-premises (was a WAN-based system so comms link reliability was the same issue either way), so they had their "cloud" tech running locally.

 

All that's really happened since is that browsers continued on their already well established journey towards platform independence nirvana with capabilities increasing to match locally installed platform-specific client-code. Oh and bandwidth costs went down and security attitudes softened quite significantly - threats haven't changed that much though, so we might yet see some raped cloud DB splattered on front pages causing a bit of a "rethink" amongst folk who likely never thought in the first place.

 

giving them the time to do the much more skillful job of making sure the data is meaningful

 

::giggle::

In the past I've had to know as much as three national (evil, kafka-esque, tagged) data format experts put together and those skills are a small fraction of the set - largely the ability to temporarily put up with concentrating on very tedious things that others much higher up on the OCD scale probably enjoy.

 

and contributing to the education of the students.

 

Which bit of their education are they equipped to help with? Consensus seems to be that school's aren't normally interested in any serious IT stuff, just the upgraded "typing" variety as the bloke with the USB-stick PC very aptly put it.

 

What the best cloud and hosted services do is to free up IT specialists from getting bogged down in keeping the infrastructure working

 

Who's bogged down in the areas you have in mind? All the DB apps in my world tend to tick over and have done for years and several versions e.g. SQL databases, Exchanges etc. Putting those in the cloud saves money for the h/w and a bit of time/expertise: a) installing, b) patching, c) health monitoring, but unless the tech concerned is stupid it's not that expensive and can be very easily be offset by new problems that cloud champions coveniently overlook.

 

I see a broad spectrum and some of the most difficult, intractable issues I've had to deal with are genuinely with "cloud" apps and they just keep on coming: A zilllion browser issues; plenty of comms & proxy & filtering & firewall issues where the default is for all the different parties responsible for any given link in the chain is to blame the other ones ('holistic' folk who understand the chain, the protocols and the effects of what they might encounter on their journey seem a bit thin on the ground).

 

--

 

Pragmatic reality for me is the "old" versus "new" tech thing is misleading and there are no universal conclusions to be drawn: A well-made local system can trump a cloud system both in terms of support overheads and flexibility (cloud may well have more constraints); the converse can be just as true; cloud app bandwidth and QoS might be an issue; native apps can still knock spots off cloud equivalents in terms of usability and features; I can often promptly fix local issues, with the cloud you might have to wait, and wait and wait; cloud obviously triumphs by delivering anywhere-access by default if that requirement is business critical and you're content re. the data security implications etc.

 

But one thing is true: Except in a strictly academic sense I couldn't care less about how a given cloud service is implemented server-side, whether it's traditional web app, TS app or [something else], it's all the various aspects of the client side experience that count and of course the ££££'s.

Edited May 21, 2011 by PiqueABoo