IrritableTech

I'd personally try to go down the radius route with a single SSID if that is possible. However Unifi now have 'Private Pre-Shared Key' which is a nice feature in theory. You can broadcast a single SSID and depending on the password supplied to connect you can tag users in a particular VLAN. https://www.unihosted.com/guides/configuring-private-pre-shared-keys-unifi-guide

Yesterday Arbor acknowledged an issue we put forward to them where incorrect school login boxes were showing up. https://status.arbor-education.com/incidents/193419 However not only did we see evidence of this issue after they had announced it was resolved. We had reports that a member of our staff had been able to access another schools Arbor site. Today Arbor have confirmed (through our support partner) that this was a caching error which did allow users to see limited data from other schools...

Seems to me like a routing issue with static.purplemash.com. DNS seems fine. I've raised a ticket with SBB for my sites this morning and I'd encourage anyone else experiencing this to do so too.

You can live feed out reports. I have some setup feeding data to Google sheets. If Arbor did go down (which we haven't experienced in the last 15 months since joining to be fair) we have the ability to view the most important data.

Glad to hear you've got it working. We've been creating a few scrips here over the last couple of months integrating systems, automating notifications, creating/editing/disabling accounts. It's become very useful very quickly to us. I wish I'd taken the time to use it well before now. Perhaps it's worthy of a dedicated thread - ideas for automations? With so many jobs and so little time, it's worth automating all those repetitive actions.

I'm not at my pc so this could be wrong but I'd get two sheets from the same spreadsheet like this. Although I'm no expert. function myFunction() { const ss1 = SpreadsheetApp.getActive().getSheetByName('sheet1'); const ss2 = SpreadsheetApp.getActive().getSheetByName('sheet2'); const data1 = ss1.getActiveRange().getValues(); const data2 = ss2.getActiveRange().getValues(); var1 = data1[0][0]; var2 = data2[0][0]; }

I don't like the look of your if statements. I'd usually be using if (var1 == 1 || var2 != 1) { do a thing } return will stop the processing.

You should try countif. =COUNTIF(A1:Z1,"Male")

We use this ourselves along with the site team, admin, finance and reprographics. Works pretty well out of the box but anything out of the ordinary usually requires you to write a plugin or hack away at the templates. The API is pretty handy.

Anyone on the beta got any comments on how well it is working? We're interested in sign in app but we'd need it to write back marks to Arbor to commit.

Redirection policies are run at logon, but Google file stream doesn't mount its drive until the user is logged on. I expect the policy is failing?

a) how did they find out - did they just ask people?? NordPass are a password management service from NordVPN. I believe like many password managers they offer a service to check your passwords for strength (although I'd suggest most users of a password manager are already above the use of 12345678), but they also 'monitor the dark web' for breaches. This is the most likely source of the information I'd imagine.

We're only half a step further along the path than you @Razzy and I'd echo what @Primus says. We've written a paper on the current situation across our MAT of seven schools. Highlighted risks associated with our networks and the additional complication of multiplying that by seven sites. We've schools with and schools without, schools teaching one thing and other teaching something else. There are two main areas which are obviously also linked. Operational and Curriculum. I wanted to make sure that operational decisions supported outstanding teaching and learning. A few points off the top of my head... Consider your core on each site, nothing can be improved without a solid foundation. Sadly it takes investment and people don't see the flashing kit, but without it the plan will likely fail. Interconnected sites need a good Internet connection, better still a more considered WAN of some type. Consider speed, reliability and redundancy. On prem, co-located or cloud. Costs, benefits, risk appetite. What you'll need where will depend on all sorts of other factors, least of all the devices users will be using. Don't forget wireless networks. They need to support many end user devices, BYOD and guests. One authentication method across the trust allows for more agile working. All devices should be monitored all the time - how can you achieve this across sites, who keeps an eye on that now you've seven sets of staff and students. How do you provision users, have you systems you can tap into to have them created when they are needed and disabled when they need to be. Be efficient. Communication, collaboration, storage. One system across the trust which ticks as many boxes for as many users as possible. Devices - who needs access to what when. Consider that devices best suited for staff may not be the same devices as those for KS1 students. Look to standardize as much as possible across the trust - it'll make it easier to support both from a technology perspective and a T&L one. How will the whole system be supported. Does it work if different providers are supporting different sites, services or devices. Does this increase or reduce risk?

We found adobe creative cloud would not login or update when the NCSI reported no internet. I’d rectify this to be honest, even if it only causes a bit of user confusion it’s a quick fix to implement.

Some good arguments in the thread but if it’s all new, will the money be available in five years time to refresh everything again? How do you spread the costs down the line, replace early, extend beyond usual life span or do without?

I’ll second (third) cloudflare. It was super useful when we had someone attempt to take down our website too. Two factor protected, multi domain, they cache your stuff world wide if you want. They’re ace.

I'm a zigbee2mqtt user, so my go to would be their compatibility list. What is exposed by each device is mentioned here: https://www.zigbee2mqtt.io/supported-devices/#v=IKEA The socket doesn't do power consumption I'm afraid. Details here on scheduling but I don't see dawn or dusk I'm afraid but that might be doable in the motion sensor because that records illuminance. https://www.ikea.com/gb/en/customer-service/product-support/smart-lighting/how-to-use-smart-lighting-pub53d86412 You just need to fire up Home assistant :-p . To be honest these days you'd have it running in twenty minutes. Integrate ikea hub in ten then half a dozen automations in another ten. The onboarding is so much nicer now than many of the guides and youtube videos show. This setup would all work offline too.

Given your requirements I’d look at Tradfri stuff from ikea (or the stuff from Lidl, but who knows when that’ll be back). They do a good range of smart sockets, bulbs, switches and sensors all through one hub and one app. They’re at the cheaper end of the market and will integrate into home assistant (which I think we all know is the direction this will end up) through ZHA or zigbee2mqtt. I’d personally recommend the zigbee route because as your device number increases it’ll keep the chatter off your WiFi network which can quickly reduce its speed. Zigbee builds a mesh network through permanently powered devices meaning your push switches and sensors powered by batteries can be further afield in your house without issue. Check but I’m fairly sure tradfri works with Amazon, google and home kit too.

As above. I don’t think this new connection has anything to do with the issue. A BT engineer isn’t going to swap cables around either. You’ve got a connection into school - your internet is working and schools broadband have confirmed this. Your physical layer is good. Now start diagnosing this issue. Are your clients getting a valid IP, are they able to ping internally? If not look to your DHCP server. If you give a client a valid manually assigned address do you get any further? Can they ping externally - 8.8.8.8 for example? If so look to dns. If not you might have an internal routing issue. Are the clients getting the right gateway address? Is the subnet correct so client can send traffic on to it? Do you use vlans? If so is your config good? Was a change made but the config wasn’t written back? Could your router/core switch just need a reboot for some unknown reason? Are clients getting a valid DNS server? Confirm it is responding correctly by doing an nslookup. Are all your dns servers giving correct responses? If not is dns running? Do you need to clear the cache? Have you actually got an issue with your AD infrastructure? If DNS is working start to investigate the schools broadband proxy which is typically transparent. Go to 888.com and see if you are getting a valid response from the proxy (should be blocked by default). If not perhaps your old smoothwall proxy has snuck back in via gpo to your browser? Have your servers and clients got the right time? Check your schools broadband proxy certificate is loaded on your clients. The schools broadband service desk can help more if none of the above helps. Ring them back, if they’ve nothing to update you with they won’t ring you back.

We paid and they migrated us. It requires them to make the API call so I’m not convinced you can do it yourself. As mentioned, most don’t charge but £150 is a tiny amount and you’re probably saving considerably more than that by moving MIS. The bit that annoyed me was sales telling us our kit was old and we could save the £150 by replacing the installation for £1500.

I believe after Google brought in the storage cap of 100TB google documents created after that limitation are included in your storage allocation. Those created previously are not.

Worth checking this list for the virtual mobile companies to confirm which network they use. Some obviously better than others in areas. https://en.m.wikipedia.org/wiki/List_of_United_Kingdom_mobile_virtual_network_operators

I’ve found ClassCharts support to be really poor recently. I couldn’t honestly recommend them as a result unfortunately. We’ve used it successfully for a few years, but after migrating from sims to arbor we have issues outstanding for three months now and zero communication.

I may be too late but after a recent experience with Zabbix I’ll ask… have you run out of space on the SD card? If you can start the pi but stop zabbix-server see what logs you can prune and free up space. That may give you enough head room to back up the MySQL database to removable storage. Reinstalling zabbix on new storage and importing the sql should have you pretty much where you were if my experience is anything to go by. All the time this also leaves your sdcard practically untouched (bar the space making) and gives you plenty of second attempts.

Totally. Its fascinating to me at least that these completely different approaches work for different people in the same industry. The question gets even more interesting when you expand it beyond a building or site, school or trust, datacentre or country...