cluberti

When a *program* is complaining about "memory", it really means Virtual Address Space, not RAM. Your dxdiag won't help much. When the problem occurs, you're likely out of virtual address space for the process, or some other internal error occurred and that's the error that bubbled up.

Not much can be done then - you have one of those rare apps (due to the DRM driver, of course) that just isn't going to run on Windows 7, and given the driver being so old, probably won't run on Vista either.

MS doesn't make drivers unless it's Microsoft-branded hardware. If your vendor doesn't have a driver listed, but WU or the Win7 CD does, the vendor submitted the drivers for inclusion (Asus has been doing this a lot lately, for instance - no driver download, but it's on WU or the CD).

Is this being run under the local admin account, or an account that is not the local admin but has admin privileges?

Was the image sysprep'ed before you captured it?

Usually you run into that when the client's check of the driver version (versus what the server has) fails, but the problem isn't the client driver, it's the print server driver (and likely it's a registry issue with said print driver). I've seen this happen many times with HP drivers on print servers (regardless of load), and the only real way I found to "fix" it is to remove the server queue, remove the print driver, re-install the queue and driver, and have clients update. Part of the reason HP moved to the new universal driver format was to fix issues like this - it's either speed or stability, your choice .

Right-click the update and select "Hide Update".

Even using Windows setup to partition/format the disks can avoid this. You only need the partition if you're on an EFI system, or if you plan on using bitlocker in the future (bitlocker will have to create a partition if the 100MB boot partition doesn't already exist).

I believe that is correct - the problem ultimately isn't the filesystem in use, it's the fact that with the filesystem code as-is, a Win9x system won't be able to access and work with files larger than 4GB.

Yes, those ARE called by explorer.exe when you right-click. Seems that the 3rd party explorer loads explorerframe.dll properly, which in turn will load shellstyle.dll and host the desktop style (classic, I'd guess) and the MUI files for it. Good catch, but explorerframe.dll does exactly what it sounds like it does, but your shell replacement obviously is coded to allow use of these binaries. Good find, but still explorer-related.

Hmmmm - what happens if you capture an image with the sysprep -pnp switch? Rebooting like that leads me to believe one or more of the drivers are corrupt, and assuming this isn't a STOP 0x7B (inaccessible boot device), it leaves the intel chipset drivers as the next likely culprit.

While a .bat file will run under cmd.exe on XP and higher (always), there are command extensions and variables that work in a .cmd and not a .bat, even when executed in cmd.exe (it is done this way on purpose for backwards compat).

If you were having HDD I/O blocking, you'd start backing up IRPs, and that video of yours would probably have hung (especially when rewinding). Given what a HDD failure or lockup does to Windows, your symptoms don't match. I'm with other posters, this is much more likely a driver issue, although I would wager it isn't a video driver issue (again, the video example seems to dispel this - you would likely have seen artifacts and ghosting). What has been installed on the system, driver and software-wise, above and beyond Windows 7?

This probably falls under rule 2.c, although the answer is in MX records. That's your hint - good luck in your studies . [Closed].

First, welcome to the forums Next, please try to post nLite-related questions in the nLite forums next time. In big red letters at the top of the XP forum where this was originally posted: [Moved].

Please note that unlike most cases where people blame bugchecks on bad memory and are incorrect, these actually *are* potential memory failure bugchecks, so you may actually have bad RAM in the system. I'd be checking that, because all of the bugcheck codes you've shown (c21a/4e) happen when a memory location lookup fails, either due to a faulty pointer or a page fault during a non-paged operation. Given it failed in Windows *and* when booting from a (very) minimal Windows PE boot disc, this is at least one scenario where I would test memory. Again, it's possible the RAM is fine and your bugchecks are just coincidence, but the likelihood of that is fairly slim.

From the Starforce page: If you've followed the instructions here and used the "5.5 or earlier" method, I'm not sure what else you can do. Supposedly 5.7 is Windows 7 (even x64) compatible, so if you're sure you updated it properly, try contacting the Star-force support folks to see if they can help. That would be my suggestion at this point.

Also, watch out for the app in question - some apps need to be licensed differently to be run on a TS (like MS Office as an example).

Just be aware that multihomed domain controllers can run into LOTS of problems: http://social.technet.microsoft.com/Forums/en-US/winserverDS/thread/a1583d7f-fa59-4497-89de-666d683e53a0 Unless your switches cannot be stacked or communicate (or you're restricting both networks for a specific reason), it might make more sense to stack or connect the two switches and use 192.168.x.x/23 instead (192.168.0.0 - 192.168.1.255, 255.255.254.0, for example). It would avoid multihoming the DC and the problems that will create.

We are all aware that the usage of programs like these for desktop OSes, like XP, Vista, Win7, etc, violates the EULA for usage of said product, right? Which is the reason why Microsoft released a specific OS version to do this, rather than something like an add-on or powertoy for existing desktop OSes. Let's try to keep the discussion within the forum rules - posting products that do this (note they're from either China or Russia, where IP and EULA laws are non-existant?) would be a violation of the EULA, which would run us afould of our forum rules a bit, and I can't endorse that.

Right, so you filter it so it doesn't apply based on the detected OS during processing.

Yes, it's a Win32 dialog box, and it's saying that it's having trouble parsing something. Not sure what you can do about it though, as I doubt IPB is heavily tested at all on Win2K or Win9x, using any browser. I'm using XP, Vista, and Win7 for my tests and none of these with FF 3.6 repro the problem you see.

Well, I think Andre is right. In looking at the CPU usage, it's definitely a consistent spike, every second - one spike is ~2.5%, the next is ~7.5%, the next 2.5%, the next 7.5%, etc. I saw what appears to be the nvidia network driver in quite a few of the 7.5% stacks, but not in the 2.5% stacks (at least not to the level down I went), but I don't necessarily think that's the root cause - although it may be, as I've seen nvidia network drivers cause this. At any rate, to get to the data analysis, the top consumer of CPU in the explorer.exe process is shlwapi!WrapperThreadProc. The shlwapi binary is basically responsible for determining if a particular API call belongs to the shell (explorer.exe), or if should be handled by the browser, and then creates the thread requested in the context required. The WrapperThreadProc being the highest CPU consumer means something *else* running in explorer.exe is trying to create and process a *new* thread, every second. Since this isn't common to explorer.exe itself, I would have to agree with Andre - what explorer shell extensions do you have loaded, and are you running the latest drivers for the nvidia devices in your system? Also, if you disable Windows Search, does the problem continue (it's a fringe chance, but it's still possible that this can cause that behavior as well).

That is indeed odd, because following those steps I get no error message with either IE8 nor FF 3.6. Hopefully you can attach the jpeg.

While it is possible someone here might be able to assist, I would strongly recommend asking this particular question over on the Drupal forums specifically: http://drupal.org/forum