PiqueABoo

Best I can do so far is to use GPO to deny 'traverse folder/execute file' for utilman.exe [Windows + R causes a "This operation has been cancelled..." for me so I must have taken that out with a GPO setting somewhere]

I'm struggling to google an effective way to turn this and the narrator bits off a) on logon desktop, b) for all but a specific group when logged on. Not least because utilman.exe runs as system. I'm increasingly tempted to just map out the Windows key, but I don't want to. Does anyone know the perfect (centrally deployable) solution for this?

How big is that VHD/VMDK? How many machines have you got? How many images can you store? And did you turn off machine password updates to save you that bother after reverting.. and is doing that good security? How fast is redirected across the network virtual disk I/O likely to be? I'm in one of those picky moods tonight sorry, but I'm struggling to see why you'd do that rather than run thin client.

Emerging.. oh joy .. yet more encouragement for folk to spend more money they don't have turning systems they only just got reasonably settled/stable into next years headache. OK that's a bit mean but there are so many technical "visions" floating about and they all seem to get automatically filed under "learning enhancement". Where's the hard evidence to show they will enhance anything besides some tech-co's profits? If this new BECTA site also manages to address that rather than simply gushing over every shiny new trend, then good.

There was for 2K -> 2k3 so there must be 2K3->2K8 upgrade paths for MCSA/Es that mostly concentrate on the new bits (and any bits they think folk are mostly rubbish at on 2K3). I'd look into something like that.

Haven't used it, but it's in system32 on 2K8 so.... ..err... no... you can use it on 2K8.

You can cause problems with a DC on real hardware. Are you saying that because of your recent USN rollback problem? I didn't really read those threads or get what had happened to cause it, but I can't think of a reason for it that isn't avoidable. Yes, in the same way it's advisable to stop as many services as you can before taking a full backup you want to restore to the same machine. Cold-cloning is the obvious extension of that precaution - if you can do it cold, then do it cold. -- More generally time can be an issue so it's worth reading about ... there's a lovely, complicated VMWare paper on that somewhere.

XP & Vista are desktop support. An MCSE is about more, but I still wouldn't care much about it compared to what the experience says. But given a pile of CVs, especially if when they're going through an HR department or some external agency, things like a tick in the MCSE box can count. [Factor in the credit crunch and maybe they'll count more - I don't know]

Around me it did a fair impression (about 20cm). OK I grew up in a part of the Pennines where they had the 8ft poles along the sides of the roads and I went out on driving lessons in worse than this, but I couldn't get anywhere with a car last time it did this down south because everyone else was in the way whenever you got to even a slight incline... ..but that's almost besides the point coz it's the most credible snow around here for about five years and you want to take you kids out to play in it at least once whilst they are still a kid.

Never needed to verify this is actually true, but Exchange heavily relies on AD[1] and I think it sticks to talking to whichever DC it settled on when the system started up. You just have to live with Exchange breaking if you take that DC down e.g. better to take Exchange down then the DC, and vice versa when starting up. [1] Exchange needed one and used to have it's own local directory. I presume (don't know) a lot of that existing tech was reused for AD.

...and no way to factor in the numbers e.g. figure out what proportion of schools are getting what amount of service from whom. Around my bit of the country it's very common for a SFE Primary to get by on 1/2 day a week of external tech support. TFEs usually get more for obvious reasons.

If that isn't small I'd get the ISP to blackhole all traffic aimed at the web-server IP for a while.. figure out whether there is anything wrong locally and if not (or when fixed) get the web server moved to another address.

I use IVT

I've got a significant NT s/w dev history and that has been my take since I spent my first day with CC4. Thus I'm not surprised to hear it, but it is interesting to have that view confirmed. You're certainly from s/w dev-land and if you're not ex-RM I'll eat my hat.

One of the key points. I don't know the precise answer, but RM said somewhere that CC4 is dual-use technology i.e. was designed for a standalone school and as part of their managed school system. If that is not the case now then it is massively over-designed.

I think you need to look at what that (internal) certificate says about revocation via view certificate or whatever W7 has now - it probably points to somewhere internal you can't get to from the outside world. You might be able to turn off revocation checking somewhere. Or you can, with a little effort, make certificates with custom revocation info.. or even no revocation info.

Well I think something is wrong. For a recent install with two, presumably test users who haven't done a lot that is much too much too big. What are the biggest files in there? When/if that's sorted, you really should move the DBs to another logical drive and the transaction logs to yet another one (and add a few hundred MB of junk files you can happily delete to make some emergency fixing space on the transaction log drive). For ref, I've got a E2K3 VM for testing with a couple of accounts and the whole thing i.e. 2K3 + E2K3 is < 4GB

DNS scavenging is a kludge, in fact it's a horrible kludge. That said it's probaby better than what most people have without it. But spend enought time around wireless laptops that sometimes use wired connections and you'll soon be agreeing with me. The main issue is the order in which DNS returns records when you've got Scavenging's no-refresh interval in action. Doing DHCP + DNS nicely is just very hard.

I love MDT. Well apart from that PE network timeout bug which took a while to track down. PXE booting is the trivial bit. I never did get 2K8 WDS to work for me, but all you need are a couple of DHCP options and any old TFTP service.

True, but.... I made a GPO that hits all users which is dedicated to allocating printers and (mapped) drives. Then you just get on and make some preferences e.g. #1 sets the default printer with a target saying "computer must be in room 27 OU". Not sure how well this scales, but it also supports what I want in places which is to give different user groups different printer sets when they're logged on to the same machine e.g. a target might say "computer must be in room 27 OU and user must be in group X"

reg.exe? That's in XP by default. Whatever, I use that. Ultimately, if it's XP native and does the trick then that's what you should use... less of anything added to a system is definitely more.

Never heard of it. Closest I can think of is a missing NTLDR... and if it's that you can make yourself an NT boot floppy to start the system.. or boot a bartPE.. and then fix it.

Yup this scenario is what makes sense to me (and most folk I've done it for on CCs etc.) a) Shutdown all workstations *without a user logged on* at say 4pm, 5pm. b) Shutdown all workstations *with or without a user logged on* at 6pm, but if someone is logged on they are prompted and can override. There's a relatively simple WMI approach to that, you just have to watch out for remotely logged on and services etc., but IME that's uncommon.

I used to spend all day in the company of earlier versions of VS and I'm happy enough with VSE 2K8. Depends what you're up to I suppose.. what doesn't VSE do for you?

Don't know whether 6u11 is any better, but the Sun 6u10 MSI had problems and unless you wrapped some magic around it, that probably hasn't uninstalled correctly. See if you've got a java quick start service running, disable that and see what happens. I wouldn't bother keeping up with these updates unless there's a pressing reason e.g. an app needs it, or there's critical security fix.