Podcast | Season 1 | Episode 18: Centralising the IT infrastructure in a Multi-Academy Trust
Andrew Walls from Sidney Stringer Academy presents the journey the trust embarked in transforming and centralising the IT infrastructure across all of its schools.
📎 Case Study: Centralised IT Network and Staffing in a MAT
- Building a centralised WAN so that you can run your MAT IT as if it is a single school
- Cost efficiencies and performance gains from pursuing a centralised MAT IT model
- Centralising staffing structures to provide clear leadership and the best service quality
- Implementing more progressive IT elements, such as G Suite and Cloud Technology, while retaining a familiar environment that makes the transition to the cloud and remote technologies easier for staff
💡 Andrew Walls, Assistant Principal (MAT IT), Sidney Stringer Academy
🏫 This session was recorded live on 14th November 2019 in the Tech Theatre of the Schools & Academies Show at the NEC in Birmingham.
So thanks for coming. The idea behind this talk is to look at how we as a MAT have managed to centralise IT. What that actually means in a day to day reality, and what that looks like for the end users on the ground. So that's what the idea of this talk is very much go through the case study of a centralisation process for it. So I'm Andrew Walls. My job is Assistant Principal for ICT across the Sidney Stringer Multi Academy Trust. We're not a very large trust. We only have five schools. Three secondaries and two primaries, but we've managed to centralise all of our IT services. So the entire trust is managed almost as if it's one single school. And this talks gonna look at how we did that. Now, the first thing a lot of people ask, and when we done this talk quite a few times for other other trusts. The first question everybody asks is, what's the point in centralisation? And you might get a few answers along the lines of it offers better economies of scale, it means that things are cheaper, it means that you have better standards.
But there's a bigger there's a more holistic approach to centralisation. I think we talk and I've seen a lot of that tech in recent years I really want to focus on for this talk. So the first thing to look at is yes, you do get economies of scale. When you make IT purchases, doing it as a large trust does make it more cost effective. It does give you more options and it does give you better opportunities to find new resources, but also expertise sharing. It's amazing seeing how much the expertise of one or two schools can then really cross fertilise between schools, not just to raise the standards of teaching and learning, but also the standards, sorry standards of IT, but also the standards of teaching and learning the standards of digital resources, student independence, there's a lot of wealth of information that can be shared on this journey. And also, the last part is it offers offers a forum for teachers to share ideas on how teaching and learning can be improved as well. So that's what I want to talk about in terms of centralisation. So just go back to the case-study, I'll talk you through our history kind of our journey that we've been on.
In September 2017. We were in a position when I came in where we had an internet connection for each school. So every school had its own internet connection, its own filtering service, its own providers in terms of monitoring and that was variable in quality between the schools. Every school had its own data centre. So our five schools you had a similar server a file server or printing server, pretty much for every skill, well, that's a lot of replication. That's a lot of servers essentially doing the same job that realistically could be centralised in that model. And also, the most costly part was everybody came into the trust with very different service and support contracts from different suppliers. And as we probably all know, on the ground, some of those suppliers were really good. And some of them not so good. So we ended up with quite variable quality that and that tends to be the case, especially when you're taking over a new school. So there was a real opportunity to rationalise and improve performance across the trust. Now, I think as well as this, the first thing to bear in mind is you might have an idea and think, well, actually, there's a really good way in which we need centralisation for these three advantages we should get this in. But also, I think it's important to look at what the consequences are of a lack of centralisation. So what we did, we did a very simple internet monitoring test for every school to check their quality of monitoring and filtering. Obviously in this forum want repeat the kind of keywords we were searching for and what kind of things we were looking for. But the more and more we try to test what we could access, we realised there was a massive safeguarding risks depending on your internet service provider, not every internet service provider by itself offers the same standards are filtering some schools with assimilated and put their own firewalls in place in order to kind of overcome the limitations of the provider. But every provider claim to offer a child friendly internet connection and quite clearly, we weren't getting that as well. This was really quite important. It might seem like a very, very kind of small area to look at. But if you want it centralisation to get onto everyone's agenda, it's very powerful to stand in front of trust directors and stand in front of senior teams on each school and be able to say, well actually, do you realise in your school, the lack of filtering and the lack of a central centralised system with high standards is allowing these materials to be accessed and filtered. It's quite a common one and we're able to make that a real priority by really scrutinise our providers when we when we form as a trust, as well as that way, every school had a data centre refresh plan, which would have cost about 20 to 70,000 pounds every 8 to 10 years. So if you add this up, and everybody roughly should refresh the data centre every decade or so maybe a little bit early if you can, that was mounted up to a huge amount of money. So if we could run a central network, where everything was just on one network with one central set of resources for everybody, the cost effectiveness just kind of spoke for itself. So it puts things on the agenda quite powerfully, really showed what we can save. And as well as this when you simulate different skills, you get very different experiences. So one school we took over had the like the slickest IT environment ever. It was very Windows based, but everything that was there would load quickly and be very instant. Other schools had very, very different policies applying that weren't as user friendly. They weren't necessarily fit for the needs of the organisation that were causing frustration. So by centralising everything, you can have a standard set of good standards that you expect, rolled out across the entire trust in terms of how you expect it to function. So I'll talk about that a little bit later on.
So you can really standardise policies, you're not firefighting between different schools. You can have your best staff who are best at creating group policies, you staff who are best at optimising networks looking at that, where everybody else can just focus on frontline support and customer service, which, which is probably about 70% of the job of it these days. So the first thing we did on our journey, as we said, well, actually, if we had to write what we expect, without using any overly technical vocabulary to make it super clear to everybody what we expect in terms of it, what should we do? And we sat down as an IT team as we we'd centralised together all the technicians across the trust. And we agreed on this set of rubrics, so it doesn't matter what's in it. It uses matters and each very end user focused, that it's very clear about what the outcome will be. We kept these conversations very deliberate, where we didn't discuss methodology. We agreed we discussed methodology later on. So we started off by saying things like, for example, all student computers within the trust, if they're running Windows, we'd expect an outstanding level of service to be every student can have a desktop within eight seconds. There's many ways to achieve that. But we said, okay, that's what we judge outstanding to me. We judge good to be 35 seconds. But anything longer than a minute requires improvement. And we did that for everything. So what would we expect Wi Fi, what an outstanding standard of Wi Fi, we'd expect at least 40 clients in a room to be able to stream high definition video without any buffering or any issues if that was required. But a good standard would be 30 clients able to stream maybe 420 or 480p. So that's where we set it was very end user focused, but that was useful for us as a team. Because then we could go and rank each school on where we were at. So it becomes very clear and very easy in a MAT structure when you need to persuade headteachers to release funding for different needs. When you can turn around and say, well, this school's got a brilliant, brand new AC11 Wi Fi solution. It's brilliant. It works. You can get up to like 1.3 gig down to a client if it's got a necessary antenna, and you can do that. But actually, this primary school, you can maybe have one laptop in the room that can maybe watch a YouTube video, if the weather's good, and everything's happy and everybody's, you know, everything's in the optimal conditions. And that's really quite powerful, because that raises the question and puts it on the agenda, or why is there a difference in service quality? And obviously, coming from an IT background? I can say, well, the answer is obvious. They've got a terrible 10 year old controller, and a terrible set of 10 year old access points, but the head teacher probably doesn't see it as a problem until you start almost I probably wouldn't say it as if they were here, playing heads off against each other to say, well actually don't you need the same standards across all schools. Don't you all want to be in the good category? Isn't that what's needed for teaching and learning?
There's many opportunities where a teacher might need to play a video to 30 devices in a room. You can think of a tonne of examples where that will be useful. Shouldn't we look at that and do it and that's quite useful for centralising and standardising. But then of course, the discussion becomes what is the methodology behind that? And from that, we can say, well, actually the best way to get some of these things standardised. If you want all the machines to log in in the same time, if you want every machine to act in exactly the same way, well, actually what we need is a centralised and standardised way of doing things. So actually, why don't we just copy the model from higher education that they've been doing for years across multiple sites, and just run the trust as one big network? So essentially, you have one windows domain for the entire trust. Every school is connected together via some methodology and we'll look at the methodology as we explored later on. And then look at what the best practices out there in large organisations with multiple sites, go visit them. And also go visit universities that have done it, because that's your due diligence. That's how you work out what the best model is. I've seen it happen in schools as well. But actually, if you want the best experience, they're the ones who've been doing it for years and years and years, who can probably offer even more expertise than most schools. There's exceptions, but I think that's a general rule.
So this is what we mapped out. And don't get too technical. So I was told not to do that for this talk. But essentially, we said, right, every school is got its own kind of split down quite small internet connection. Why don't we just buy a nice, fast gigabit symmetrical line for the trust? We've got five schools, that's probably about right. But why don't we go with a provider, which can scale that up, and we can get more speed. In the event that we take over more schools, or more schools asked to join. Because really, with a MAT, you've got to think about future proofing in terms of new schools coming on board and that could be anything extra two secondaries, it could be an extra primary, you've got to be flexible. But essentially, this Trust has 5000 live devices, on average, it does go up and down at any one time. So that's what we had to accommodate for that could expand to 10,000. It's, you know, reasonably quite short notice potentially. We then use the filtering argument to make sure that we had a decent firewall in place or a device that could monitor and filter the internet connection. And I would say, the key argument for me when we were putting this in wasn't that it offered a better standard of filtering across the trust. But it meant that all internet breaches, all incidents will logged on one firewall, just a single box that can be checked every day. So you haven't got five people in five different schools, checking the firewall logs at the end of each day, or if I'm honest, because some of the schools were so small, probably not checking the logs at the end of each day. Probably check it once a week, or once every two weeks, which isn't Good enough. If you think about small primary staff are overstretched, who's checking keyword searches going through that firewall for things like suicide risks and self harm is the somebody who can do that every day in a trust there is I can do that I can sit there as the director of IT, and then sit and look at the central set of logs at the end of each day as a job. So for safeguarding much better. Then we planned to link all the schools up, and that's where the due diligence came in. We contacted lots of providers who go out there and provide wide area networks for trusts that will claim to offer links between schools, we contacted about 12 different providers to look at this. And a lot of the options are very different. So you can and one thing we explored first, and I'll just make sure it's on as it is an extending screen, was to look at this website which is Solwise. And what this does is it shows you line of sight links between points. So that you could, in theory, connect high capacity, what's known as mini wave dishes to wirelessly link two schools together. So you link two schools together over an incredibly high speed wireless link. We looked at this, and you put your postcode in and you put one site in. So if I put in one school and then I say, well, could I also connect to another school in our trust, and then you just tick the links, click where you want to look. It's really easy to do.
But if I just wanted to click on two links together, so I know this is an extreme example of a distance. It's not feasible. No one operates at that far. But if I just wanted to, I can see where the obstacles are, where the issues are we potentially looking at line of sight. So that could be that topology doesn't allow for it, or that it's just not feasible to do. So if I just pick a link there that could quite feasibly in that link, you can see that there's an issue with terrain there. But actually, if we're talking about linking these two sites, there's a small gap there. But if we try to link a school there and there, oh, look, it's a great example here. You can see there's blockers in terms of line of sight. This is a ground level, if you're doing it, to look at the height of buildings you can put in that it's actually 10 metres above ground level. And let's say that this one here is 20 metres above ground level. Put them in together you can see all well actually if you get a dish maybe a bit high, we could get a link. So that's what you've got to play around with a very kind of practical example of what it looks like. But that's what we looked at first, we looked at a wireless option. The more we looked into that, the more we realised Coventry didn't have this or enter to really support that. And when you look at building plans, it wasn't suitable. So he then in turn contacted five or seven fibre optic providers who could provide links between schools and in the end, resellers for city fibre were some of the best. So we ended up creating city fibre links using provider providers who could work on city fibres backbone who run fibre optic and, and dark fibre that's being put in the ground that is just completely never used. So it's you sitting there waiting for someone to connect buildings up to each other, but essentially never is connected up to each other, make it live use it. And it's a lot cheaper that point, your single internet line can be shared between different schools, your filtering can be shared, and everybody saving money because they're not all spending on an internet connection.
So that's what we looked at. We ended up going through a central wide connection. We then started looking at other solutions we could use to make things easier for staff as well. And we implemented Google Cloud Print throughout the trust as a kind of early stage of getting people involved in in more progressive it elements in using G Suite and tools that are quite progressive. And when we start using this, this was great, you could print from home very easily completely for free, you could print from any device. So that eliminated a lot of issues and got people kind of buying into slightly more progressive IT but allowed us to share good practice and set things up that we wouldn't otherwise have been able to do. And then later on, we ended up with a central data centre, a single data centre, where all the different schools could set share everything together. So one file server, one print server, make sure that everything was optimised in the same way on this network and then slowly moved all the schools onto that system. As we moved everybody onto the system. Everything became a lot lot faster, a lot easier to use, because we have the best policies, we have the best setup and everybody's running the school all the schools are brought in like that one single school working together. So that was what we ended up doing all the quiet a different period of time. So that's that's kind of the situation we're at now. All the schools are sharing one internet connection, all the skills of sharing a single firewall better for safeguarding monitoring and filtering, but also one data centre, which means that everything is stored in one place. And that allows us to make sure we have a consistent backup policy. So we can put a backup server in a completely separate school. So should God forbid one school burned down, the data is still safe there. But also then, in turn a third layer of backup outside the city, where somebody else through our providers can back our data up as well. So we have our data stored in three locations to provide complete geo redundancy. If you're a small primary school, these options aren't available to you. But in a large central MAT structure, it suddenly just becomes a part of the ecosystem.
Essentially, a primary school is just like an organisational unit or a very small part of one larger network. So they get the same treatment as everybody else. So they get the same exact setup. The consequence for this is the standards are better. So if we look at our policies, this is the slowest machine we have in the trust, which is an old HP 5800. It's about 15-16 years old. It's for the scrap heap, but essentially, the state we've got it is it will log in in 13 seconds flat, because we've been able to optimise those settings and get everything exactly right. And that is only available in a large MAT. If you're a small primary, your only option for having a setup like this is to hire consultants, experts in group policy writing, it takes a long time, but actually, if you look at secondaries that are established, we have that expertise to be able to share so we can do that. And we can set it up and make sure that everybody's at the same standard. In terms of providing equal quality of service elsewhere, we can offer services that other schools perhaps couldn't afford. So that centralised data centre offers a remote desktop service for each school. Not only do we offer remote desktop service, but we have because we have the technical expertise of secondaries with lots of experience of doing this right for large members of staff at large numbers of members of staff. We can set up a remote desktop service that will just run in a way browser, so you don't even need to instal anything. And this came out with Windows Server 2016. But it's since been built into 19 as well, where essentially you could log on to a school computer or remote desktop environment with lots of redundancy, lots of backup, lots of potential for growth from wherever you are, whenever you need to, and that'll load very quickly. And that service is identical, whether you're a large secondary, or a small primary.
Just as I wrap up towards the end of the time. Some of the other things that we've done as well, is what I said on the teaching and learning. So I won't go into all the teacher learning things we've done because time doesn't doesn't permit for that. But one thing we did set up and it did allow us to as a MAT is we populated google classroom, for every school in the trust as well. So there's a single platform for being able to set homework, share work, record work, submit work that every school can do. Now, that's a free tool. Anyone can use it, but in terms of our expertise, we were able to publish this on my blog as well. We were able to create a method which we blogged about where we could keep it synching with SIMS. So every school had Google Classroom setup. So it linked to a SIM. So the submission of work was identical now, and it's really fun to work with easy teachers, can we classes, they essentially had a free VLE that didn't cost any money. The small primaries of course, we're paying for the VLE's or the schools within the trust, we're paying for VLE's quite expensive, whereas with this set up, and the expertise we were able to pull together, we were able to set up essentially a VLE structure using Google Classroom. Because we were a centralised MAT, we had the expertise in one location using a free tool. And because we had the technical skill to do it, making it sync with SIMS so people actually gained functionality they didn't lose it. So we've done a lot of different work in terms of our our use of technology and being able to share it.
Another thing we rolled out well, we didn't just focus on the server side, as I said, we focused very much on the cloud side as well, we gave every member of staff a Chromebook as their default working device as part of this move as well, within the trust, but what we did is we gave them the option, you could work in G Suite, we could push that as a very 24/7 you can work from anywhere cloud based environment to work in. But also, if you needed to, you could go to that remote desktop, that familiar Windows environment that many people needed in order to feel comfortable. So we didn't force anybody at some schools do, you must go down the cloud route, or you must go down the remote desktop route, we gave staff the choice. It's like, Look, you can use the progressive G Suite way of working, or you can use the remote desktop to use things you're familiar with. And that way you don't lose functionality for things like SIMS for tools that perhaps aren't quite web ready yet, and to and different software that just isn't a cloud version of the software. Yeah. So that's kind of our journey in terms of centralising IT. Don't treat it just as a technical process. It's an expertise sharing process. And it's also one where it's worth making sure that you put a strategy in place that works for every school level that maybe by themselves they couldn't afford. Okay, thank you.
Don't forget to register for your free place at our upcoming show on www.schoolsandacademiesshow.co.uk.