AlanD

Wait. you must NOT use channels 1,5,9 and 13. I don't know who or when the Wiki article was written - but its poor advice. it MIGHT have been possible in the early days of "g" wireless to get away with squashing in 4 channels, but that was because it did not use orthagonal whatever mejig so the usage of the 20MHz channel width was very low at the edges of the spectrum of the channel. Even since "n" came out you get much more complete use of the 20 MHz channel - including at the edges. - and spacing 1,5,9 and 13 will produce overlap and co-channel interference. There have been some learned and impressive mathematical papers produces by Cisco amongst others on using 4 channels and they all agree that the co-channel interference using 1,5,9 and 13 will more than destroy any advantages of the extra channel. So Sorry - you are stuck with 1,6 and 11. you MIGHT....if your school is isolated and has no neighbouring housing go with channels 1,7, and 13 which would give you better channel spacing - but you would need to be absolutely certain no one is going to use 1, 6 or 11 anywhere on site (or on a peer to peer link to a projector or whatever). So no. Absolutely no. You must not use 1,5,9, and 13. No vendor/supplier should advise that.

In fairness - I will point out that the MERU single channel architecture is not without some merit. Its often dismissed - in disbelief - that a single channel solution might offer better throughput than conventional wireless deployment...but in dense wireless environments it can be effective...and you can still layer additional channels if it bothers you that you are not using all the spectrum. So if you have 30 laptops all streaming Video then expect most premium wireless solutions to start buffering and spluttering long before a MERU system. And it "may" offer a way of using wide 80 or 160 MHz channels - although in reality there is little feed back from schools to suggest that there is any benefit with "ac"...and I have always wondered if the chips sets don't actually allow MERU to do what they did with earlier "n" chip sets. which may have lost them that edge in the market. And I agree - the interface is awful. A true example of how "NOT" to build a user interface. I still drool over the meraki interface everytime I see it until looking at the price gives a reality check. You may find a supplier that is happy to lend you Unifi kit - but given that they are so cheap and the profit margin not nearly so lucrative you will probably need to buy one. You could buy a wave 1 AP just to get an idea .....and probably sell it on ebay for the same price as you paid for it if you didn't want to keep it. Of course in luxury market suppliers will beat a path to your door and knock the door down if they think they are going to hook you into annual contracts.

Unless you have some very old "g" kit all access points will do "MIMO" - which is essentially what "n" access points brought in. Well worth considering the Unifi HD model....which is 4x4 and supports MU-MIMO (which is perhaps what you refer to). MU-MIMO will give you increased benefit to even single stream wave 2 devices - and there are lots of devices - including laptops which are ONLY single stream. The access points you refer to are not wave 2. For a new installation I would be trying to go down the wave 2 route because it does have some benefits - although I agree the access points are not nearly so cheap (..although still a lot cheaper than wave 2 kit from the luxury market) Think very carefully if you are replacing "n" wireless kit with "ac" and don't get rosy eyed by all those gigabit figures you see quoted such as 800Mb/s or better. To start with its for UDP traffic - and you will what TCP - so divide it by 2. Then its for 160MHz channel - and there are no current access points that support 160MHz channels except for home routers...and even if there are some you can't use 160MHz channels in dense classroom environments only isolated sports halls or something. So divide by 2 again. In fact - you probably can't even use 80MHz channels sensibly - so divide by 2 again. By now you are essentially down to the 40MHz channels that "n" offers. MU-MIMO will give you gains with suitable clients of something like 1.2 (Access points vendors often quote things like theoretically up to 1.8 ...but its nice theory....in the real world its just not like that even in ideal circumstances). Putting access points in corridors outside classrooms completely does away with all those high speeds through walls....in fact 5GHz doesn't even go well through bodies - so you need ceiling mounted kit in each (and every) classroom if you are actually going to get high speeds. I despair at the ignorance and waste when some schools spend fortunes on new wireless kit and then deprive themselves of any benefit because they either can't be bothered to run cables to the centre of ceilings of classrooms or have no budget left to do that. ...and of course...without suitable clients the latest and even most expensive kit is not going to deliver anything...and sadly few laptops come with 2 stream and 3 wireless "ac" laptops are almost unheard of. And this is because they produce them largely for a home or sparse business environment where they can use the wide channels that ac offers. In reality they are not really suitable for schools with dense wireless environments. Vendors will extol the virtues of channel sharing and quicker on boarding and offloading of new devices which give older devices more air time...which is true...but unless you get your stop watch out with some pretty artificial tests you are not going to notice.

In short No. I don't think you should take away printers from the classroom. I think teachers are generally busy - really busy and are not "free" to wander off to get an extra print out - or even perhaps an occasional forgotten class set of print outs. Anything that can be done to make it easier and more convenient for them should be met within reason. Monochrome laser printers are really cheap to run - and often run well on "compatible" toners - we don't find toner sticks quite as well to the paper in some cases. It also provides some "redundancy" when the main printers/copiers break down. Now I'm all for reducing costs and questioning whether this or that really needed printing, but I don't think it is the role of any IT manager to take away the facility. It might be the role of the headmaster or SLT...but not of the ICT team. I particularly dislike offers from printing companies to "reduce your printing costs"....when what they really mean is "reduce your printing". Typically if a teacher requires and will make good use of a 30 worksheets it is not for us to remove that facility or make it more difficult for them by making them walk halfway around the school and back to save 10 pence on that print job. I know printing adds up....and there lies the difficulty....but a teacher's time costs money too. By all means encourage teaching staff to use the most economic print method and ask them to print class sets in advance (and not to include unnecessary additional copies in that print run). Help them by offering to do the printing/copying for them - especially if it can be printed as an A5 booklet stapled rather than A4 sheets...or get the secretary to offer this service for them. ...but if they then need an extra 2 worksheets at the start of the class because someone messed up - don't take away their printer to prevent them. Its a good thing they can print what they want there and then without wasting their time. By all means "log" printing using printer credits but don't use that as a means to restrict whet they need - because that would be hindering the education they are providing. By all means encourage - and if necessary twist their arm to be be organised to do the printing in advance... But honestly....the difference betwenn monochrome cost per page on a printer in the classroom and a large central printer cannot make much difference and certainly not worth the inconvenience to staff. I think this is a case of ICT staff getting above themselves and forgetting they are in the school to provide a service. Yes, an economic service and a cost effective service - but not to make decisions as to what staff can can can't do. Risograph printers (essentially inkjets) are probably the cheapest way of printing - possibly less than half othe last of running a laser - BUT - the quality is medicore compared to a laser and the initial purchase cost is eye watering. In what way are classroom printers a nightmare to manage? I find a litte education goes a long way. Often teachers try to add paper while printing is taking place....that is always a mistake and will cause a printer jam. Show them that removing the existing paper in the middle of a print run does not break anything, add more paper and return the tray and printing continues. Show them how to lift the toner out to easily remove a paper jam. Warn them not to go tearing off jammed paper - and not to poke a screwdriver into the fuser roller. And allow them to print to classroom next to them so that when there is a jam they are happy to wat for you to clear it later. ...Really shocked at you thinking it would be OK to remove these printers....sounds like a real asset.

I think there may some confusion here. I think the original question is about using a SINGLE fibre (rather than a pair) for a duplex connection. I don't think its about Single Mode and Multimode fibre which is do with scattering of light and multipath reflections over longer distances. I've never heard of creating a duplex connection over a single fibre - but it doesn't surprise me to hear that it might be possible. I assume that there are specialist fibre modules to do this - although I suspect range is reduced because there will be "noise" scattered back. Maybe it would get you a working duplex connection if one fibre failed...but suspect that in most cases the cost of two such modules would be a lot more than dragging in a new fibre run and terminating it.

...Wondering if this is because support was provided indirectly through local authority....then its become an Academy ...but no one thought through ICT implications.... So I assume...there have been no regular software updates to your CC4 either....and there have been lots of "critical" updates that ensure the smooth and trouble free running of CC4 in the last couple of years. I am beginning to have some sympathy for RM - which often (and sometimes deservedly) come in for a lot of criticism...but if this is from schools that have chosen to abandon support ...and yet clearly are not so averse to using it that they have gone "veneer" (possibly because there are some good things about RM?...) then perhaps I am not surprised. I think there are better "tools" - which can "virtualise" an existing server which are more tolerant of different hardware the recovery takes place to. But yes, in theory at least - you should be able to recover the backup to a virtual server. I can't see the logic in you wanting to "test" an installation which is bound to be problematic with missing updates.... If I thought for moment that it was an attempt to - for example - to clone an RM server for use in another school - then that would be unacceptable, unethical, almost certainly illegal....and while a Judge might have some sympathy with an excuse of "lack of budget" I could practically guarantee it wouldn't reduce any sentence handed out. As someone who has written and published software I expect users to pay for it. If they don't want to pay for it...don't use it. While I question the "worth" of RM CC4 regularly (and indeed the worth of Microsoft and Apple software) they are entitled to charge what they like - even if we don't like it.

That's often what I hear from subject teachers. It may even be true (Although actually there is a lot evidence that it can promote learning. Whether its "Can promote" or "Does promote" usually depends on the teacher )...but my job is to promote the learning of ICT skills....and you won't learn those unless demands are being made which stretch what students (and staff) do. And that is much more likely to happen if there is easy and natural access to ICT hardware. Seeing and using ICT as part of everything that is done seems much more appropriate than hauling a class off for some contrived and artificial activity in an ICT suite.

First - and most importantly - this is something you should discus with RM support. You pay a small fortune for it - so make that money earn its keep. And surely you are already doing full backups....what is different about this backup from those you already do? My understanding is that (assuming all updates and patches are in place and the CC4 server is running OK) is that CC4 does its own "backup" of PostgresSQL. I'm not sure how often it happens - and probably its just the one backup file which gets rotated . I've never had to use it to recover- and I guess it would require some kind of manual intervention (such as stopping the service and renaming files). You could always start the server in "safe mode" - so there are no services running and perform a backup to a local USB3 SATA disk or something - which might be safer because there could be other services with "open files" and cached data not actually written to a file. I am wondering why you are asking this...and what you intend to do with this backup. While bare metal recovery might be possible from a windwows backup ...I'm not entirely convinced I'd want to put it to the test unless I really had to. If one of our servers went down - I'd more more inclined to take the opportunity to do a completely clean re-install and then recover the user/shared data only. It would seem too good an opportunity to miss. A load of stuff (including AD) can get very touchy if you present an existing network and with a "recovered" server.

OK - not familiar with Learnpad ... perhaps its worth the £1000 (?) you pay every year for Unify..... I wasn't aware that RM safety net requires RM unify to work....(which makes the cost of RM internet not so attractive...) RM books...how much is this used? Do they use it with tablets in the classroom? ...RM support.....(so you don't have to type your logon details in....that is not worth the cost of unify...) Adding your own shortcuts.....well athere are mllions of ways of providing links that students can access - including creating tiles in Office 365. You certainly don't need to develop your own solutions..... there are ready made ones provided for Office 365 and Google. Download install and configure (just like you - or someone did to set up Unify). I'm not saying it not worth anything....or isn't "handy"....its just that I think schools pay a lot for it - largely on the basis that it is potentially an "enabler" but then don'tgo on to make use of that - largely because the packages are n['t there. ...there are also other solutions like Mgaellan for half the money which does the same....

Such as...Kerboodle...(No)....MyMaths (No)....Blippit (no) ...Flat.io (no) .... So none of the really good useful ones, just a load of old ones that no one uses - or some trivial new ones that cost a fortune and do very little..... Sorry getting increasingly cynical and grumpy ... I see schools investing in RM unify - to support all those packages that seemingly offer so much ...and then rarely does any school buy into any of the packages - and effectively only use Unify for G suite and O365...and they could have had that integration for free (and then complain they are short of funds...). Which packages do you pay for that work with RM unify?

What does RM Unify provide you with that GADs and GAPs do (for free)? such as..provisioning from AD...and password sync....

Don't need a network to do a windows backup - to a USB3 SATA drive. So I would be doing a basic backup even if core network went down....but its up to you.

...I've always though backups were the Number One priority....To do them, check they are done, and test they work. Refuse to do anything else until that can be ticked off as job done weekly/daily.

That would be an over simplification. I agree you only have a set bandwidth...nominally 1Gb/s and dividing that up into vlans doesn't help deliver any more bandwidth to voip...although you might be able to throttle bandwidth on the domain LAN with some switches....The main improvement is to remove all the noisy clutter that bonjour and other broadcast services are creating which voip devices would hear and decode even if they then decided the traffic was not for them.

OK, so its hosted. What happens to voice calls when they reach your core switch? Do they get passed separately to your router....does your router maintain that QoS? Are you using a leased line (or FTTC or cable modem?) Do you reserve bandwidth for voice traffic - or rely on QoS? Think I might be asking the ISP for names of other schools businesses that use their Voice service - to contact them just to double and triple check this is not something at their end. Wouldn't take their word for it. ...might be worth updating any firmware in Cisco...routers...printer firmware....printer drivers.... (Why only 3 phones on this VLAN rather than all of them? Is that because it was working before...with without that...and so just a test?) Is your wireless on the same native VLAN as phones? Or as the computers? Sorry - I'm not sure answers will solve the problem - but just trying to get a picture of what you have - and your reasons for doing it that way. I hear the statement....it used to work.....so its tempting to think its something you have added/changed or updated....but it could be something further upstream with your ISP despite their denials....hence needing to find other customers that use their Voice service.

Sorry not read the whole thread.... Are these drop outs between internal calls - or are they only for calls which are external? I can see you have been advised to use a VLAN for your phones. I would agree with that as a starting point. And then make sure QoS is enabled at every switch for that VLAN. And make sure you don't have any loops (or possibly a loop created using wireless). If you have any wireless...I'd be tempted to turn that off for a day to see if that is related to the problem. My guess is it will be because of things like tablets and printers using "Bonjour" to discover themselves in the background - which can produce an enormous clutter of traffic....but once you have created the voice VLAN that will disappear. Getting someone in isn't going to have any magic answers or tools....you can do what they are going to do. Turn bits off (starting with wireless) until you have narrowed down the issue. Yes it will be disruptive. They will do this too. Its possible that even "internal" calls are actually going via your external internet - do you host your own VoIP exchange internally? So my second guess is its your internet router/filter/connection that some issues....or even an issue with whoever is hosting your VoIP.

My vote would be for at least CAT6A Yes the cable cost is higher....but not much higher and the labour costs must be about the same. No point in paying the same money again in 10 Year times (a lot can change in 10 years....10 years ago we had some 10Mb/s coax networking ). You might easily want better than 1Gb/s to wireless access points (although I don't think any yet support 10Gb/s yet) ....

Extended service support for 2003 R2 ended in 2015. It would be difficult to justify (...and please don't "cost" as a justification) for keeping with 2003 R2. With successful ransomware attacks not uncommon in schools, and a requirement to meet Data protection criteria (if hosing SQL on it) you could be accused of a dereliction of duty and negligence (probably a sackable offence) for not taking reasonable precautions - and that would include using "supported" versions of software in my book.

Ubiquiti do their nano range of point to point links. Is it pout to point..or point to multi point https://www.broadbandbuyer.com/products/28801-ubiquiti-lbe-5ac-gen2-ptp/ There is a guide to point to point referenced towards bottom of this web page.

Capped at 2mb/s ??? You have a business rather than residential service? This is fttc not adsl? We have plus net and get a steady 20mb/s upload on any of the speed tests and between 76 and 80mb/s down ...but rarely use anything above 5 mb/s up. Never had any experience of a cap for speed or useage quota.

I think the lower front end cost of S/H PCs often blind schools into thinking they are getting better value for money. And it often makes me smile to read on these pages of schools complaining about poor performance....especially of laptops....and on other pages of buying second hand kit. I suspect a new PC lasts for longer than 5 years - possibly 8 years or more in many schools - which makes the S/H value not nearly so attractive as it might seem at first. I'm not against buying secondhand stuff - I buy lots of it...but for secondhand PCs I would want to pay a rock bottom price and install an SSD. Usually we "gut" old computers and install a new motherboard/CPU (and get a new version of Windows of course). Some of our "boxes" have been through 3 generations of CPU. Wish I had hung onto some of the old keyboards and mice - though - because the current ones seem very flimsy and fragile.

Yes - I rarely notice our servers capable of saturating 1Gb/s nevermind 10Gbs....the exception is backups when I assume the sequential data can be accessed and moved much quicker that the random access in typical use. Collectively our servers might make good use of the 10Gb/s - but That is atypical - except for copying and backing up data when its good to know that you are not throttling the network for other users.

Perhaps I'm not understanding something.... We have several old optical fibre runs....really old...probably even before there was a classification of OM1, OM2....multimode 65u. The runs are over 100m long ...maybe even 200m ..... and we just stuck a 10GB module (Multimode) at each end - bought for £10 each from china - and were pleasantly surprised that the connection worked just fine - with no errors showing up in switch statistics. I am at a loss to understand why you need special patch cables - unless it is to go from 50u to 65u diameter ... we find that cheap optical patch leads of either diameter work just fine. We did get the old fibres reterminated some years ago as a job lot when some new fibre went using replacing the old home made glue and crimp connections with spliced end pieces - which obviously reduced potential loss.. Our switches are netgear..built up over the years for £50 each buying them off ebay so we have been able to move to a 10G backbone for under £2000 - including a spare couple of switches in the cupboard - so we think we have done OK. Each cabinet has 2 x 10G switches with at least 2 x LACP fibres (one from each switch) going to edge switches (we are a physically large campus site). Any switch in the backbone can fail without affecting connections to edge (apart from a potential drop in bandwidth). Cabinets have UPS (again bought cheaply off ebay with new compatible batteries installed)