AlanD

I've ben looking into this - for the same reason - lack of space for T: and W: drives.....and yes - it would appear to be a difficult and tangled knot to move them elsewhere - and even then it would seem to be necessary to use another DC. So, I started looking into increasing the disk space. Our CC4 servers have two mirrored system drives and four RAID 5 Data drives which fill the 6 slots. But it appears to be possible to add a 5inch to 2x 2.5 convertor and add another two drives into the RAID in our rack case and use OnLineCapacityExpansion to rebuild the RAID without losing data. RM wouldn't support this - because we don't have a hardware contract....and I am not enthusiastic about mixing two new drives with old drives. So I will probably replace all the drives. Obviously - good backups required first...probably stop all RM services before making backup and restart them after data is recovered.

Keep in mind that "filtering" is not the main requirement of the prevent strategy....but monitoring is. So I would be looking carefully for a "filter" that reports on the categories mentioned in the prevent strategy - such as self harm, radicalisation, etc - and you need reports on searches and sites visited matching these - without necessarily blocking - otherwise you never get a chance to monitor if there is cause for concern. At one time the philosophy was to "block" all sites which might have safeguarding issues - even if there might be a good educational reason for using them (use of drugs ...for example)...but that has now changed.

Sorry, I'm not familiar with the HP range. Do I understand you have used up all 6 slots in the HP switch? (I assume the "06" means 6 modules). And have you used up all the existing fibre and copper ports? I was partly being devil's advocate - but, also making the serious point that if there is a problem somewhere with "speed" - you need to understand what that restriction is (if indeed it is the problem) - so you need to look at and monitor network traffic. While you may have 10G connections to your servers - are you using a SAN? And if so - what SAN switches do you have? Are they 10G too? Do you have a SAN capable of delivering multiple GB of data? Perhaps you do. Perhaps you have SSDs with hot and cold storage, or NVMe mirrored drives in your server boxes - in which case you probably are getting several GB of data into your core switch. Although I would still come back to created aggregated (Trunked) links from your core to edge switches before I started looking at replacing edge switches for 10GB ones....not least of all because it can be done for free - or nearly free. Yes we do use some 10G links - but rarely are edge switches pushed above 1GB/s - mainly because our servers struggle to get the data out that quickly. BYOD - especially if you are feeding it with a 500MB/s leased line or something can consume network bandwidth of course. We use "cheap" netgear switches - bought second hand off Ebay a few years ago for £100 each - with 10G modules (another £25 off ebay ) plugged in. I'm not suggesting that everyone would want that kind of solution - but it is possible to provide performance on a shoe string. (We have a couple of spare netgear switches sitting on the shelf just in case the main ones fail - which at that price you can afford to do of course). And yes....the fans are irritatingly noisy in netgear equipment.

How do you know its creaking? How do you know its not your servers? Or the links to your edge switches? Or indeed the 100Mb/s network interfaces in your PCs. No point in in holding your finger in the air and thinking "perhaps I'll upgrade the switches" - because that might not be where the problems lies...assuming there is a problem. The danger is you could - as you describe- invest a considerable sum in 10G switches and no one will notice a difference - which would be a complete waste. In the first instance - I would be putting LAGs between core and edge switches....not least of all because it provides you with redundancy...and next, because it costs almost nothing....other than some configuration and would almost certainly provide you with the added bandwidth your require. . To fill a 10GB link anywhere probably requires a server farm and a massive SAN. I bet you have at least two cores of fibres (I never put in less than 8) running from core to edge.... and if buying fibre modules for your switches - use Gbics.com or try some of the £10 ones from China - they seem to work fine - well better than fine actually.

Aerohive - controllerless - fascinating. I thought I was the only one in the country to do this...although...when I say controllerless - I use One access point in a each building running version 3.5 of the firmware to "manage" the others. (I think you can use later versions of the firmware and enable Web interface via command line). I do use command line stuff for odds and ends too. Paid £10 per access point from an Hotel chain getting rid of them. Does radius authentication via smoothwall. Aerohive firmware is essentially autonomous in that it understands neighbouring accesspoints and arbitrates and negotiates between themselves so apart from configuration (which can be done via command line) they don't actually require any central management controller. They work really well - but I guess using them this way would not be to everyone's taste.

I think that I am right in saying that Ubiquity is widely used in schools...and integrates just fine with AD/BYOD etc. Has no on going annual costs and probably less than half or even quarter of the cost of some wireless kit - a veritable market disrupter. Sorry - can't comment on Huawei, because I've never tried it. But I would put in a mention for Xclaim access points which made by Aruba - and again have no on going annual costs and come at a budget price for a serious piece of kit (limited to 4x SSIDs) with on site management tools. It only supports two streams but that's as much as most devices manage these days. Clearly Aruba brought these out to compete with the likes of Ubiquiti which robs the premier market of their customers. But be aware that Aruba has just been taken over....which might be fine...or not. Are these budget cost access points any worse - well possibly - slightly - but you'll need to do a load of tests to verify it and the difference will not be noticeable.

You need to consider the prevent strategy, where the emphasis is on monitoring, not necessarily filtering. So...can you"see" what students are typing into searches? Does it create alerts wherein these searches suggest radicalisation, violence, self harm etc. Does it produce reports categorised by those things mentioned in prevent.....and not necessarily block them...lists of 100 most blocked sites isn't much help. And of course you need to identify this by user...which may not be easy or convenient in a primary school...especially if using tablets.

I've looked at RM unify several times....but we already run Azure AD connect (to sync Office 365) and GAPS/GADS to sync google accounts....and we have created some Javascript/ASP to log students into some sites transparently...so I think its quite a lot of money to spend every year for a few (albeit convenient) tiles and not having to enter your password again to get to a site. I wasn't sure RM were actually syncing with MIS yet...but they have talked about it. Not sure how it will cope with different MIS systems - I guess it might only work with their own Integres. And yes - it would be great to export users and groups in real times and update timetables or calendars automatically...but its not there - or at least not yet. You could create tiles and menus within Office365 to produce a similar look to RM unify....for no cost. I don't think you will ever see the equivalent of RM's OneDrive mapper (or any other such product) for google because it doesn't/can't use WebDav - and so it always going to work by using Sync methods - which would bring any network to a stand still.

I've looked at RM unify several times....but we already run Azure AD connect (to sync Office 365) and GAPS/GADS to sync google accounts....and we have created some Javascript/ASP to log students into some sites transparently...so I think its quite a lot of money to spend every year for a few (albeit convenient) tiles and not having to enter your password again to get to a site. I wasn't sure RM were actually syncing with MIS yet...but they have talked about it. Not sure how it will cope with different MIS systems - I guess it might only work with their own Integres. And yes - it would be great to export users and groups in real times and update timetables or calendars automatically...but its not there - or at least not yet. You could create tiles and menus within Office365 to produce a similar look to RM unify....for no cost. I don't think you will ever see the equivalent of RM's OneDrive mapper (or any other such product) for google because it doesn't/can't use WebDav - and so it always going to work by using Sync methods - which would bring any network to a stand still.

My understanding was that Google Drive - if mapped by whatever means - is always going to cache and sync stuff locally. It cannot make use of WebDav technology that RM's OneDrive Mapper and IAM essentially make use of. ...but I'll be very happy if someone can convince me otherwise.

My understanding was that Google Drive - if mapped by whatever means - is always going to cache and sync stuff locally. It cannot make use of WebDav technology that RM's OneDrive Mapper and IAM essentially make use of. ...but I'll be very happy if someone can convince me otherwise.

I am a relatively new user to smoothwall - and so still getting the hang of it. I thought the reporting looked fine - no worse than Websense (now forcepoint) that I was used to. And I liked the Safeguarding provision a lot. But...the safeguarding reports don't have the customisation that the main reports have (didn't realise that earlier ).....so I would quite like to email reports to appropriate Heads of year which ONLY have the reports for the users in that year group. Don't understand why safegiarding section doesn't have the customisation tools available elsewhere. ...and I'd like to be able to clear up some of the "clutter" to make the reports more understandable/readable without the multiple line quoted content. And currently I'm forever having to "unblcok" or allow sites - such as downloads from exam boards - which are clearly going to be safe for schools.... ...oh...and APPS on mobile devices are a bit of a nightmare - I seem to have to practically allow a list of URLS for each APP to bypass https inspection or authentication to allow them to work. But I'm not saying the competition is any better....because I looked at several of those. And I liked the fact that BYOD works nicely with it using it as a Radius server and DHCP server....and I like the reverse proxy....and the load balancing....

Slightly concerned that you are confused about budgets. Budgets do not mean that you must spend them. Indeed some of the time you might need to spend more - because of unexpected expansion, extra kit, etc - which is counterbalanced against the fact that sometimes you find you can spend less. I've had laptops expected to last 4 years fall apart after 2 years, and desktop PCs last 7 years that we originally thought would not last 5. Senior Management are very happy when you say I am going to delay upgrading this or that because actually it works fine for now. So unless you are absolutely clear what improvements you want to gain - do not simply replace Access points (or anything else) because some arbitrary timescale is up. You get no brownie points for that. You might well find - for example - that user experience would be better spending money on laptops (or tablets) which had network cards that supported 3 streams rather than 2 (and were ac ready) than changing the access points which might bring no improvement at all to existing users. I'm glad to know you have a number of SSIDs - and run separate Vlans - becuase that is the way I would advise anyone to run things. (Keep in mind that too many SSIDs are not a good idea because they take take to advertise themselves away from actually transferring data). i'd look at the speeds that devices which don't have an AP in connect at. (there are lots of mobile apps or speed testers). If you are not getting good 300Mb/s on an "an - 5Ghz " connection I would make it a priority to get 1AP per classroom.... Integration with AD is easy if you use smoothwall - just make the access point a radius client from smoothwall - which then handles the authentication from AD which I am guessing its already connected to. If you are going to install wave 2 "ac" access points - bear in mind that you may require 2 x CAT6 cables to each access point. And you might require "at" capable PoE switches (or injectors) - or else sacrifice some of the benefits of the that new ac" AP. Some manufacturers claim their AP will run off "af" PoE - and fail to mention it will be only with one of the two raiods working, half the streams, or without the second cat6 cable connected, or without that BlueTooth beacon working. ...but I will keep coming back to ..."YOU" need to test replacement Access points. Personally - as far as functionality and performance is concerned - I don't think there is much to choose (having been through this myself) between the premier league players - although I do like the Meraki interface a lot...and I do think that the MERU single channel architecture can deliver amazing results provided the collision domain does not have too many access points (i think there are some horror stories when single buildings have tens of access points..) Ruckus will deliver a better signal in a weak area - but getting 2Mb/s rather than 1Mb/s isn't going to solve the problem of not having 1 AP in each class room. We use Aerohive - although we manage them through the command line because we are far to poor to pay their cloud management costs, But whatever you test you do don't omit the Ubiquiti ones - although you might struggle to get a company to supply and set them up - because they are essentially designed for you to do yourself. I think I would start with buying a single or set of Ubiquiti APs for a block of well used classrooms. Make the SSIDs the same and tie them in through smoothwall for AD authentication. If you decide later you don't need them they are easy to sell on ebay for near the cost you pay for them. Use them as a benchmark (a) against your existing kit and (b) against any supplier trying to sell you new kit. Don't get taken in with a single AP test with it set up for a 160MHz channel - because you are not going to be able to use that channel width when you have a cluster of them. You might to use 80Mhz channels, especially if you can use DFS channels, in which case - and providing your end user devices support it - you will see an improvement over your existing "n" APs, but I bet there is not any noticeable user experience difference between the Ubiquiti and permier league stuff....which is kind of not that surprising because they use the same chip sets...and use the same basic microcode in the firmware supplied by the chip makers.

Have generally avoided managing mobile devices until now (except ipads - for which we use Apple's own profile Manager and MDM stuff) Thought I'd use Google G suite - but the Android devices are Kit Kat - and that is apparently not supported. Looked at Spice works - but the "free" stuff doesn't seem to do the useful stuff.....which seems to be true for other free offerings. Perhaps we just need to bite the bullet and pay for Meraki MDM?

First you need to ask what it is you want to achieve. Your devices already does "n"...and you are probably saying to yourself that you want "ac". Any why? Because you are wanting and expecting faster networking, higher capacity/bandwidth etc. Now - I'm not saying that you won't achieve that - but its not going to be as revolutionary as you might imagine. "ac" is only going to be faster (a) when you use wide channels - perhaps 80MHz or even 160MHz and (b) by using more streams © MU-MIMO . Cynically - I might point out that (a) you can't have wider channels because you have already have high density of access points so classrooms to the left, right above or below you are going to generate a lot of co-channel interference. (b) More streams is not going to work - because even some of your new tablets and laptops don't do more that 2 streams - and probably don't use the 3 streams you have now (are they 3 stream APs?). © And your devices generally don't support this either - apart from maybe BYOD mobile phones. OK, Yes - I know the MERU single channel architecture is going to allow you to try to operate all the access points on a single wide channel. But there are lots of cases where schools are finding they need to abandon this - because using separate layers (or channels) each of which are small channels - tends to deliver a more satisfactory result to classrooms in close proximity. If you are not already 1AP per classroom - I would work towards that first....possibly by keeping/moving the MERU kit to one part of the school and buying new kit for the other half. 5GHz does not work well through most walls - so it doesn't matter what new kit you buy - unless the AP is the classroom you are going to lose a very significant part of your bandwidth. The days of needing wireless surveys are long gone. Unless you have a shed load of money (and it sounds as if you might - because you are asking for a supplier to install, setup and test......its not exactly rocket science to slap a few access points up) I'd be taking a close look at Ubquiti - and buying a handful of Access points to test. Cheap as chips, and no annual charges and they do all the essentials like band steering (if you must) and Air time fairness, etc. Does the premium kit from Meraki, Aruba, Ruckus, Aerohive and Uncle Tom Cobley and all do better? Yes - but you would need to do some accurate timings to know this - and no user would be able to tell which you were using unless you told them. Don't take my word for any of this. ...Don't take anyone else's word either - especially not anyone trying to sell you their kit. Get some test kit in. Most are happy to supply test kit for free (Meraki will give you an AP just for doing this, so with Aerohive....and others will loan kit to you). Be prepared to be blown away by how neat Meraki's interface behaves - but sit down before you look at the pricing because its going to shock you. Everything should work with smoothwall - which I think is brilliant with BYOD because you can run it as a DHCP and RDIUS server so users sign in once with their device - and that's it. No captive portal stuff or anything to worry about.

Its not that £16 is ridiculously cheap - its that £240 is ridiculously expensive. Pricing today seems to be based to "what will people pay" - rather than what does it cost to produce and market - whether its airlines seats, mobile phone contracts, or household energy. I guess for many businesses ICT infrastructure is peanuts - and so they are happy to pay these prices. In schools - our whole budgets are less than their petit cash spends.

I'd like to add my vote of confidence in compatible mods too. Had no end of trouble with genuine Netgear ones costing hundreds of pounds each failing - so eventually bought some at £10 each from china. Had to wait a month for them to arrive - but they have been brilliant - had 10Gb and 1Gb modules - all great - even on long and age old fibre runs...and unlike the genuine ones - they never seem to fail. I've also bought some from Gbics.com - not quite as cheap - but worked brilliantly too.

Well - I certainly agree with the conclusion. Assess the products out there and evaluate their strengths and weaknesses. It would seem to me that too many schools take a look at what others around are using - and use that as an argument for implementing it. I thought SW did better than most to address the prevent strategy....and provided a hardware firewall with the ability to provide relatively seamless BYOD use as it acts as a radius and DHCP server. And we were keen on a solution which could do reverse proxy. Yes it has limitations,....and I am currently struggling to get lots of APPS to work for mobile devices without either making them "invisible" by bypassing https inspection for them, or blocking them - or risking holes that allow users to bypass authentication. Its good to prevent users from being able to download potentially infected word macro files, but simply blocking all word files is not so useful. And it would have been nice if files from "official" sites like exam boards and government were not blocked by default. Perhaps they should take a look at the "exceptions" that lots of schools end up adding and after inspecting them (and making any appropriate changes) have a "school configuration" that is likely to meet the needs of schools without lots of school based customisations.

Not much point in focussing on remote cabling if the server structure itself is not going to deliver anything. A RAID with 10 drives is going to struggle to fill a gigabit link and indeed might only have gigabit links to your main switch...hopefully some link agrregagated ones.... And massive disk RAIDs, even those with SSDs are not always much better as they uses SAS expanders which limit overall data rates. Only if you have NVMe drives in the server and the server has 10G link to main switch would I start worrying about what was further away. Yes, aggregating some 1Gb links to main switch would be a useful addition. Running video editing to lots of stations via a network is of course questionable as a concept...and I doubt that is going to work very well without serious investment throughout. Better to have a load of local storage to work with as temporary space. We have old pre om1 fibre which runs just fine at 10gb...albeit over a distance of 100m or less. And we use aggregated links throughout between core switches and perimeter, for redundancy if not bandwidth. Yes we have some mono mode high grade fibre for long runs. We think 1gb to desktop makes a differnce for user experience even though our servers would struggle collectively to deliver this to multiple stations sat a time. ...and of course wireless and ever increasing external bandwidth, schools I know of gave 1gb leased lines, need bandwdth in the infrastructure to deliver it to ac access points.

Overloading certainly seems to fit the facts.....so I'm happy to back that idea up.

...will be interested to know how you get on with this. Why not simply migrate to o365? What do you gain by going hybrid? Because...as I understand it, once you choose hybrid you must keep your local exchange server to manage mail..even if there are no local mailboxes. To address your question..once the hybrid is set up are you not looking for an option to move the mailbox?

Well...I would say this is for Meru and its suppliers to sort...but I'm guessing its going to be up to you to get to the bottom of what is wrong before they get involved. I've always found that sitting in the chair and saying - well it worked fine yesterday - doesn't get anything fixed. Its true, that sometimes after making a load of changes ending up with everything as it was when it started mysteriously leaves you with everything working fine. Make a list of the things it could be - and what you can do to check. One room, One AP (everything else OFF). Everything OK? 4 rooms, x AP Working now? Make sure everything else is off (no interference from a failing AP or something). Make sure no one has a rogue AP which is active. Why no Try band steering off. Somewhere, you will get to point which helps you pinpoint the problem.

I'd be really wary of band steering - especially if you have every other room where 5Ghz coverage is not great. If you know you have a strong signal everywhere - its usually fine. Anyhow- you need to work your way methodically through trying to isolate the problem....even if you think you have not changed anything...perhaps something you changed that didn't take effect before has now kicked in (and band steering would be an easy one to eliminate).

Do you have a Volume Licencing Agreement for Office - and does this allow your students to have Office 365 accounts? Its the setup of the administration that will prove to be tedious - keeping track of parents email addresses etc. There has been talk of systems that will tie the school MIS system to Office 365 (for Google and Office 365) - which I think is common in the states - but MIS systems are varied here - so I'm not sure how that might work for everyone. It will help if you are a wiz are sharepoint - but yes in theory it could be done. In practice - I'd want a year of 6 month sabbatical to set it up.

OK, so you need to eliminate the MERU access points. get something cheap and cheerful off ebay, buy a an ubiquiti AP (and sell it again afterwards)....borrow one...bring your broadband router from home and use it as an access point - just to make check to see its your network or the wireless. If its the wireless - then MERU needs to sort it - which they should be really keen to do - because I bet you pay a fortune annually to them and you won't be paying them after you realise that it might be cheaper to swap all your access points for "ac" ubquiti ones and save yourself those annual charges.