cluberti

You're doing just fine - and making users administrators is sometimes the only option for traveling users (especially sales personnel). I'm always against it if you can avoid it, and there are always ways to do software installs/management via SMS or Group Policy, but traveling users are challenging to any administrator.

For a small organization, maybe. But Exchange is such a resource-intensive application that for anything else other than a small install, Exchange on a DC is just downright dangerous. Both Exchange's store.exe and the DC's lsass.exe functions are *very* resource-intensive, and having them on the same box can run you out of resources - not to mention you have to use /3GB for Exchange, which means you already have 50% less nonpaged pool, and almost that much taken from paged pool (not to mention about 180,000 fewer PTEs to start with on boot). Don't do it if you can avoid it. Seriously.

If you're on a domain, you don't need a .net account, and Remote Assistance can be configured in Group Policy as well. RA is completely domain-aware.

What happens if she tries to launch IE7 with the -extoff option? (C:\Program Files\Internet Explorer\iexplore.exe -extoff)? This is almost always caused by an add-in, not IE itself (an add-in modifies the URL passed via wininet or winhttp, and IE then tries to open something incorrectly - you can also see this a lot of times when IE adds an extra www up front, or tld extension after, what you've already typed).

Right. 9x is perfect - except for a lack of multiprocessor support (and thus a lack of SMP support), lack of ACL support in the file systems nor the registry, nor any at the process level (thus user-mode applications can crash the whole system), lack of user permission levels, and only 64K of GDI and user-mode (desktop) heap compared to 3MB (default) in an NT-based OS (meaning I can't have as many applications open at once). I won't go into group policy and domain support, as that doesn't affect home users as much, and multi-user support in XP/Vista probably isn't as important to some home users as well.

Can the discs be read in another machine entirely? It seems odd that both your Mac and your PC are having trouble reading discs (format shouldn't matter if it's supported by both, as UDF should be) - so what is in common amongst these discs you've burned? It certainly isn't the OS, so is it the application used to burn the discs? Is it the drive on the machine burning the discs?

First, in the future, please post the actual file, not the contents when there's such a large # of lines of data Next, Gosh is right - when Windows hangs at the splash screen, it's usually NOT a driver loading, but a service load that's at the heart of this issue. Does the same thing occur if you boot into Safe Mode?

Generally, this error is very bad as it means that one (or more) of the registry hiv files on the laptop disk are corrupt, and Windows cannot boot if one or more are corrupt. You may be able to get lucky and be able to boot with the "Last Known Good" F8 boot option, but if not, you'll be reinstalling Windows if you don't have a repair disc handy: http://support.microsoft.com/?kbid=282501&sd=RMVP

It would be even better to configure it for a complete memory dump while you set it to not automatically reboot (it's in the same location as the "Automatically reboot" checkbox - just make sure the min/max values for your paging file are configured to be at least the size of the RAM in the box, and you'll have a .dmp that can be analyzed as well, above and beyond the bugcheck code.

Unless you point a fan at it

All time in the OS is done in UTC, and then the offset for the time zone is used when displaying time or timestamping. And everyone is right, in an AD domain time sync should be done to the PDCE, or the closest DC in your site, whichever is closer. You shouldn't have to do anything other than make sure the PDCE for the domain has a valid time source to sync against.

Which NIC did you enable sharing on?

What happens if you run word via the "Run as Administrator" right-click option?

SBS doesn't come with any - you either install Certificate Services on the SBS server and create your own (untrusted outside your domain) certificate, or purchase one from a vendor like Verisign or GoDaddy. Yes.

Let us know if this works for you or not.

You can try using the subst command to map a network path to a drive letter location, but I can't guarantee that will do what you are looking to do. However, there's really no other way I'm aware of with built-in tools (perhaps there's a 3rd party app that can do this, but someone else would have to provide info on one if it does exist, as I do not know of one).

You may want to take a look at the following articles: http://technet.microsoft.com/en-us/exchange/bb288524.aspx http://www.pocketpcmag.com/_archives/Apr06/directions.aspx

You also need to consider whether speed in write is important, or only read speed. RAID5 is great for read speed, but somewhat poor in write speed. If your users will be writing files to the array regularly, you need to determine whether or not speed is important to your user base doing the writes (and what kind of writes, large files or small). RAID10 is much faster than RAID5, but at the cost of more disk space to parity (a RAID5 of 4x 300GB disks gives ~900GB, whereas a RAID10 of the same gives ~600GB of storage).

The only other wrench in the works is the use of security on the switches, like 802.1x or other certificate/radius implementations, which can slow the process down immensely. Although, if this was the case, you'd see it on desktops too, likely.

It would be interesting to see a network trace from the client port as well as the DHCP server port during this time - does the time delay seem to occur on the client, or does the initial discover packet get to the DHCP server and then it takes ~20 seconds to send out the offer? Where in the DORA process is it slow, and on which end?

Fizban pointed you to the article, but the error you are getting is "ERROR_WINHTTP_INVALID_SERVER_RESPONSE". Since your proxy requires auth before you can access the internet, it is likely that this is where the error lies, and the proxy may need to be modified to allow access to Windows Update without auth.

It would help if you could provide even a minidump of this (although a complete dump is infinitely more useful). I'm expecting the module was win32k.sys, Atapi.sys, or Ataport.sys, but if you could post that as well, that'd be great. Note that a STOP 0x7A means that a page (4Kb) of kernel data was requested from the paging file to map back into RAM, and it could not be read into memory - this generally means bad RAM (the page was corrupt when written to the disk from RAM earlier, or the write operation from the disk to the RAM location failed) or a bad hard disk controller or a bad hard disk (failed to read the section of the pagefile on disk where the page to be remapped into RAM was supposed to be located).

Looking around the internet, I found this thread which seems to match your symptoms exactly: http://www.pcguide.com/vb/showthread.php?t=42196 After looking at the dumps, I can say that it is either a problem with the Nvidia driver itself, or an interaction between the BitComet software and the driver (as per the thread above). Uninstalling the Nvidia network driver and _then_ disabling the device is probably a much better approach to this problem if upgrading the Nvidia driver doesn't fix it. They're all the same (the addresses are usually a little different, but the thread stack and exception record are always the same), so here's the analysis of the first minidump: EXCEPTION_RECORD: f8157acc -- (.exr fffffffff8157acc) ExceptionAddress: f511c5c8 (tcpip!TcpipBufferVirtualAddress+0x00000008) ExceptionCode: c0000005 (Access violation) ExceptionFlags: 00000000 NumberParameters: 2 Parameter[0]: 00000000 Parameter[1]: 01000006 Attempt to read from address 01000006 kd> .cxr 0xfffffffff81577c8 eax=01000000 ebx=00007ed3 ecx=1eb4ade0 edx=00000000 esi=01000000 edi=00000000 eip=f511c5c8 esp=f8157b94 ebp=f8157b94 iopl=0 nv up ei pl zr na pe nc cs=0008 ss=0010 ds=0023 es=0023 fs=0030 gs=0000 efl=00010246 tcpip!TcpipBufferVirtualAddress+0x8: f511c5c8 f6400605 test byte ptr [eax+6],5 ds:0023:01000006=?? kd> k ChildEBP RetAddr f8157b94 f511f27f tcpip!TcpipBufferVirtualAddress+0x8 f8157bb4 f5120b32 tcpip!XsumSendChain+0x44 f8157c34 f512094a tcpip!UDPSend+0x3ca f8157c58 f51209b0 tcpip!TdiSendDatagram+0xd5 f8157c90 f511f308 tcpip!UDPSendDatagram+0x4f f8157cac 804e37f7 tcpip!TCPDispatchInternalDeviceControl+0xff f8157d14 804e37f7 nt!IopfCallDriver+0x31 f8157dac 8057d0f1 nt!IopfCallDriver+0x31 f8157ddc 804f827a nt!PspSystemThreadStartup+0x34 00000000 00000000 nt!KiThreadStartup+0x16 kd> dps f511f27f f511f27f 840fc085 f511f283 00016e4b f511f287 031476ff f511f28b ff5350c7 f511f28f 1585ac15 f511f293 8bd88bf5 f511f297 368b1446 f511f29b f73bff33 nvmcp+0x43f33 f511f29f 7d39ce75 f511f2a3 30850ffc f511f2a7 6600016e f511f2ab 5e5fc38b f511f2af 08c2c95b f511f2b3 01f98300 f511f2b7 4916850f f511f2bb 4f8a0000 f511f2bf 07f98001 f511f2c3 a0cb840f f511f2c7 f9800000 f511f2cb 80337409 f511f2cf 840f0af9 f511f2d3 000241d3 f511f2d7 0f0bf980 f511f2db 00927085 f511f2df e8565700 f511f2e3 000439c7 f511f2e7 6683f88b f511f2eb 02b2001c f511f2ef 7e89ce8b f511f2f3 6015ff18 f511f2f7 8bf51581 f511f2fb feeee9c7 kd> lmvm nltdi start end module name f5106000 f5118e00 nltdi T (no symbols) Loaded symbol image file: nltdi.sys Image path: nltdi.sys Image name: nltdi.sys Timestamp: Mon Apr 23 07:02:47 2007 (462C9257) CheckSum: 00015DEE ImageSize: 00012E00 Translations: 0000.04b0 0000.04e0 0409.04b0 0409.04e0 kd> lmvm nvmcp start end module name f737c000 f7461000 nvmcp T (no symbols) Loaded symbol image file: nvmcp.sys Image path: nvmcp.sys Image name: nvmcp.sys Timestamp: Wed Apr 13 15:34:21 2005 (425D743D) CheckSum: 000E8E41 ImageSize: 000E5000 Translations: 0000.04b0 0000.04e0 0409.04b0 0409.04e0 In the future, kernel or complete dumps are easier to debug!!!!

Are these wireless or wired connections?

Hmmm - normally I'd say generate a dump of this behavior, but since this is almost always caused by antivirus or firewall software, try uninstalling the antivirus (and 3rd party firewall software, if it's installed) and reboot to see if this changes the behavior.