Michael

Of course other considerations with Lease Lines and Point to Point, is they can be used for multiple services, such as VOIP, CCTV, Door Systems, so they all eat up bandwidth. If you add up all the line rentals coming into your building per year, you'll see it makes sense to channel that money elsewhere.

I have to agree with Colly - you won't get the same 99.999% Service Level and with a lot of services going cloud, your upload will make a big difference. Of course the other advantage of a lease line is that you can rapidly increase the speed to 500Mb/s+ very easily. Using any form of copper - BT Infinity or other and you're stuffed.

Implementing 10Gb in core infrastructures isn't that uncommon, but 10Gb to end points in your LAN or 10Gb Point to Point across a WAN are also possible, but I've never seen it. Cost obviously plays a big part and the simplification of having one connection instead of bonded connections is obvious. Arguably if you're running everything at 1000Base-T and your switches are all connected at the same, then potentially they can or will at some point become the bottleneck, just as it's happened with 10/100 networks - implementing gig backbones did make a noticeable improvement. In the context of a school, the speed of network connections are probably more important than say a business; simply because you have 30 people all logging on, accessing shares and accessing the web at the same time, all trying to achieve the same result. This doesn't happen in a business environment as typically everyone works independently. You also have to consider the speed of hard drives in workstations - typically a hard drive will do anywhere in the region of 25MB/s > 60MB/s or more with an SSD, yet to max out a 1000Base-T connection consistently is considerably more difficult.

Are these two domains/servers on the same LAN or separated by a WAN? You could try disabling the software and set a bypass (on each IP), if there are any hardware firewalls in place.

Normally (top of my head), it'll disable the old computer object in AD and create a new object, but I'm not 100% sure top of my head if this is still the case?

I've not used LANSchool, but use NetSupportDNA in plenty of schools - best advice I can offer is to try their 30 day free trial and go from there.

Theoretically yes, but the best advice is to take a full backup before you start.

I can confirm that's the exact behaviour I've seen too - this is why I think my theory holds some water; broken Start Menu and corrupt Apps, but absolutely no idea why it happens. Microsoft clearly have their head in the sand on this issue unfortunately!

I respect all you guys/girls trying to get this working reliably/consistently. Keep up the good work as one of us will no doubt solve it. The Start Menu no show issue appears on all builds to date after a time, so I can only conclude it's profile, App corruption or possibly even registry related, as others have pointed out a re-image is the only solution to resolve it. This is why (at the time), with some reluctance I went down the Classic Shell route and I very rarely have issues with users navigating the Start Menu. Because it's a separate application to the core OS, even on workstations where Apps become corrupt, the Start Menu (Classic Shell), is still functional. I'm of the opinion this is why the results are incredibly varied, as other XML options (such as controlling Explorer file extensions), works consistently/reliably, so the only difference between this and the Start Menu is the Apps themselves. The XML file's pointing to numerous Apps and if these Apps are corrupt/missing, it may explain the random behaviour of Windows 10's Start Menu.

Would one of these work? They have both 10Base2 and 10BaseT connectors.

As for Feature Upgrades - I'm on the fence, again it depends how big they are and what it will offer to your end users. But if you're running 1511 and say Redstone 2 is 1704 (as a complete guess), you would most likely notice the improvements. 1607 is faster than 1507 and 1511.

And to answer your other query, always image. It's actually quicker than in place upgrades and leaves no old system files behind.

I don't think it's unreasonable to skip versions - For example, I deployed 1507, skipped 1511 and deployed 1607. Most likely I will skip Redstone 2. I guess it depends what exactly new (if anything) it will bring to the table.

I presume you've configured DNS? This is crucial so they can properly resolve each other. Creating a Stub Zone on each domain is sufficient, as it creates a read-only copy of each other's zones if that makes sense.

Well same rule applies really - if the Staff you support would benefit accessing such a database outside of work, then it does make sense.

Still makes no sense (depending on the application), as only a minimum of bandwidth would be needed to relay info to a website, whereas all your users are in house. I can only speculate as you haven't said what the database is for. If the database needs to be accessed outside of school or it would help Staff work more efficiently, then yes I'm all for it. If it's something they'd never use outside of school, then there's little benefit in hosting in the cloud.

A gigabit connection internally (on your LAN) should be 1ms, but a 1Gb WAN connection may not necessarily perform the same, as outside your network it's a completely different ball game!

Forgive me for asking, but why couldn't you have allocated a public IP to your server hosting your database in house, then create a rule, so said IP can only communicate with your website's IP for security purposes? You have to be careful as the 15ms response time may not be your connection. If their hosting provider is only using a 50Mb connection, the bottleneck is their end (if that makes sense)?

10ms is very fast I agree. You'd typically get 1ms on gigabit connections, but of course this will change as soon as it's being properly utilised.

The plot thickens... I've even gone as far as setting Chrome to be the default browser (using a GPO), then deliberately setting IE as default, rebooting and my XML file is working, as all defaults are restored to Chrome. I've tried the same steps on my own notebook (not domain connected) and the same behaviour occurs in Word 2016 clicking a hyperlink. The issue might be linked to 'Always use this app' but not entirely sure if I can supress this setting or not!? Can someone else replicate this behaviour please?

Hi all, I have a Hyperlink problem on Windows 10 Edu 1607 I hope someone can shed some light on - Using a GPO, I've created an XML answer file so Chrome has all its defaults. I've even created a custom value, so .URL also opens with Chrome, yet when I click a Hyperlink in MS Word, the end user is still prompted to choose a browser. I'm 100% certain it's working, as I can verify this by looking at Default Programs. Not sure what else Hyperlinks on a document would be classified as extension wise!? Many thanks!

On a related note (to the above), I've added associations for http, https, htm and html so they open in Chrome. This works for links (within Explorer), but if a user clicks a Hyperlink in Word, it still prompts the user. Any idea what I need to add to stop it prompting the end user so they open in Chrome automatically? Thank you!

I've had nothing but trouble with Edge (as of late), but only with interactive type PDFs - Chrome was one solution, but this caused issues for some users too. I used the following to control Explorer file associations (via GPO) and it works a treat:

That makes some sense, but I hope you see where I'm going with this. I've seen it happen with Macs, only to realise most of the applications are Windows only. Yes you could Bootcamp or virtualise, but it takes a great deal of time and training.

Other considerations are what will actually run on a Chromebook? For example last I checked there are no Smart Notebook or Promethean solutions and there are many other 'Windows only' solutions, especially in Education. No doubt this'll change in future, but it's certainly worth thinking about.