PiqueABoo

Dunno about yours but we're usually too busy sticking pins into effigies of school NetMans..

Al you can really do is verify that the reg value LoginMode=2 (where "[blah]" is probably "MSSQL.1") in this key: HKLM\SOFTWARE\Microsoft\Microsoft SQL Server\[blah]\MSSQLServer Then wait for Capita to sort it out (what do you think your LEA can do?).

Thanks, but none of those work with my constraints.. multiple sites, want to minimise change & don't want to add nix boxes to them.. there are strict ACLs for outbound traffic that I can't change e.g. can't get syslog traffic out.. would like to avoid SPFs. Would have made do with HealthMonitor but when testing memory usage climbed to 170MB in a couple of hours of not doing very much Looks like it might have to be d-i-y dotNet code or maybe hack the c++ Snare agent. Neither are trivial.. catching events isn't too bad until you consider filtering, suppression of duplicate events, finding events that happened say during boot before your service got going etc.

I want something free that monitors event log events on a box and sends them somewhere in e-mails. HealthMonitor is OK, but the filters aren't so fabulous (you can't make an filter for a specific combination of ID + Source). EventSentry Light has very flexible filters but being free it deliberately doesn't monitor the Directory Service log etc. Any others?

Not entirely sure what you're doing but Exmerge has worked for me (saves mailboxes to PST file that can be imported on another Exchange box). So what's wrong with PRFs then?

You should NOT do that with E2K3 because it breaks. I expect the same is true for 2K7. Look at PRFs if the concern is to get clients hooked up to the new server.

Is that true for all workstations or just some? It's not something irritating like case-sensitivity is it? Does specifying width&padding for the number make any difference? Well I wouldn't - it's not exactly easy to type "ping ws005056e28cfc27", convey the name via a phone etc. (I'm assuming that's the kind of name it is supposed to make)

Yes, because it's still Windows and you can do almost everything anyone can on Vanilla. But in practice you don't need to because the off-the-shelf config of the desktops, accounts, GPOs, server(s) file-system organisation and permissions et al do the job. There's very rarely any need for a script, but they still made a GPO for that just in case. You might need to change the odd setting in a GPO e.g. screen lock time, or stuff you're supposed to config e.g. proxy settings.. but that amounts to relatively minor tuning. Well if you're reasonably clueful, organised your systems in a sensible way, already fixed the inevitable oversights in initial design, have obtained tools or written thoroughly debugged scripts for this that and t'other.. then it probably won't. In fact it could make things harder: RMs output is imperfect and you might not have made their mistakes.

Errm... no, but I'm not making it up.. although my comment on this was too throwaway and a bit crap... I think VMWare uses AWE. That allows a process to allocate lots of nonpaged physical memory to put things in. To access that physical memory you need to write code to map it into into your 32-bit virtual address space. With that bit of d-i-y memory management you can use more physical memory than will fit in virtual memory. Don't know what the overheads are like generally or in any given implementation, but note you avoid some that normal userland virtual memory has (e.g. pages on disk). Certainly when I read about VMWare using AWE wherever it was a few years ago (probably their forums) they didn't suggest you used the /3GB switch... so I just assumed that adding another GB of virtual address space makes no difference to their memory management.

Ermm.. no. For instance VMWare have long said you can make a 3.6GB VMWare guest on a 4GB box without the /3GB switch (because it doesn't map a VMs memory into user address space).

People don't post about sucess. But that's typically because CC3 appeals to monkeys[1] and a lot of them aren't very clueful about drivers i.e. the idea that they might have to get some "custom" drivers for their exotic hardware. Even when they do understand that part, they often don't realise that maintaining a comprehensive library for every gizmo on the planet is a non-trivial task, not least because of the poor quality of some of the driver packages, in particular their INFs. The BTS forums bear witness this (driverpacks.net). [1] Not the same thing as "all people with or all people who like CC3, are monkeys" Another is you get me working for you for free Seriously... I often pinpoint/workaround bugs in CC3 and RM may comprehend and patch them for everyone one day. Trouble is that also means I'm working for RM for free Sure and if people really are good at Windows they'll cope with CC3 in about the same time it takes them to cope with any random Vanilla system and much faster the next time because it's a standard config. Vanilla isn't a standard and there's no telling how long the next one of those you meet will take to decrypt. See 4 below. One? What about: 4) All of the above.

The MS line on this kind of scenario was to give [sensitive group] their own member server(s) for their top-secret data & apps and require IPSec on those servers.

Naah.. thats a tech with a pile of CDs spending a day in front of each PC. So what is wrong with RMs? I can't see any major difference between that and any uhh.. classic method. (Comparisons with relatively recent WDS/AIK stuff don't count).

Chances are they already bought OpenView etc. If you have the money in the real-world you buy tools to make management easier. Native Windows management still isn't that good, especially for larger orgs.

This kind of problem is typically because something installed on that box has damaged bits of the very fragile driver signing "ecology". Look at KB822798. The catroot2 rename in Method 3 often solves one of the most prevalent causes of non-admin penstick trouble - it's harmless and worth the two mins it takes to try. [Caution: I'd skip Method 9]

I know a couple of what I thought were obvious things you can't do on an CC3 network, but they're part of my Vanilla B*llsh*t Detector[tm]... as in the day some critic mentions one I'll know they know enough about vanilla Windows for their comments on the particular RM config to be taken seriously. I don't think CC3 makes any particularly impressive Windows skills redundant, but most CC3 people have 3rd party support for everything remotely challenging, didn't have and don't acquire such skills. But if they don't outsource all that then they're dealing with pretty much the same stuff as everyone else.. plus all the spanners RM throw at the system. Lots of different jobs out there. If you go work for any serious real world org they will have a well-formed network already, with management tools, standards and procedures. If it's an international org you might find the latter get set by some team in another country. If you can coexist with CC3 chances are you won't have trouble adapting to someone else's existing "restrictive" environment. See also: Any BSF thread for the techie playground vs. corporate management debate.

I wonder if credential caching makes the difference? Whatever, cleaning up the associated reg keys is neater. All you need then is some elusive magic to make any orphaned "hsperf" files and the like go away - things with nasty NTFS permissions requiring take ownership, files with illegal names etc.

Would have thought we'd have heard by now given the recent noise about unsanctioned WU updates to WU itself. Perhaps it's some malware turning the firewall on to prevent the competition (other malware) from trashing it's new home.

Just stumbled over this site (looking for something else) and haven't got a clue if it's been remarked on before, but http://www.arctech.co.uk say: Doesn't say, but I assume that's per seat?

MS downloads has scripten updates for 2k, xp and 2k3 (with newer versions of some of the 5.7 DLLs you get with IE7). Don't know whether it does anything we care about.

Perhaps they're set to share less memory. I've encountered 256MB boxes with the BIOS configured to use 64MB for graphics.

The NSA, or at least one bit of it, was "reaching out" almost a decade ago e.g. a bunch of us were asked to scrutinise and comment on their draft unclassified Windows NT Security Guidelines (two of us did - both obliging Brits - and the rest including a few US citizens apparently didn't).

I've imaged boxes to USB HDD using imagex via BartPE... and it will work the other way around.

Never tried it, but I'd be surprised if the ISA Web Proxy will route to a different upstream proxy based on client user/group . Best you can do might be students use the Web Proxy chained to LEA #1, staff use SecureNAT to get to their proxy LEA #2.. which may be what you have already.

Done & posted in Downloads forum.