cluberti

According to the Lenovo support page on that model, it only shipped USB 1.1 sockets ( http://www-307.ibm.com/pc/support/site.wss/MIGR-39624.html ):

Unfortunately for you, nvidia actually makes drivers for Vista that are worse than other vendors like ATI (and that's saying something).

If it's always when using java-based webchat, consider uninstalling the currently installed JRE from your system and upgrading to the latest version from http://java.sun.com...

What controller is your boot drive attached to? The 7B stop is common around here when the driver for your RAID or SATA controller isn't integrated properly (STOP 0x7B = Inaccessible boot device).

Well, since we know now it's not monitor .inf-related, I'd say yes to that.

Understood - but what would happen if you changed the monitor's type in device manager to Generic PnP as a test? I've seen Dell monitor drivers cause the same issues (even though the .inf specs state a higher res), whereas the Generic PnP works just fine.

When you're under the advanced settings, under "Monitor type", does it say "Samsung SynchMaster" or "Generic PnP Monitor"?

Well, the event log is a good place to start - run eventvwr.msc to see it. Check the system log specifically.

The error is coming from an INT call to the BIOS for processor functionality, and a response is being generated that your processor speeds are being limited from full capacity by the motherboard - either for thermal reasons, or because there's some other issue and the processors are faulting. It's time to contact the motherboard vendor and see if there's a BIOS update or firmware update that might address this...

Hmmm - safe mode bluescreens??????? Ouch. Might be time to scan that box for malware, or perhaps even better, rebuild from a clean source...

OK, so something installed on the system (with what appears to potentially have global hooks, if it's running in that many processes) is trying to do something it cannot, causing a priv exception. If you run autoruns and shellexview and disable all non-Microsoft items (and reboot after), does the problem continue? Also, does it occur in safe mode?

No, it won't affect DNS or your AD. The only thing it would affect related to your AD would be if you installed a certificate services server on the DC (don't!), as that relies on IIS. Assuming it was a default install of 2003, and cert services isn't a default installed component, you don't have to worry about this at all.

These are your options. Note the most important part (in your case) from the article:

You're failing with the error "DSGetDCName failed with 59" every 10 minutes (Entering with timeout 600000 and flags 0x0). So, the client cannot contact a DC, thus causing the error every 10 minutes. If you can fix that, you will likely clear up the userenv error.

Hmmm. Hate to state the obvious, but I'd plug in directly to the internet source and try it to make sure it wasn't a browser/TCP stack issue. Assuming that works, it would seem as if packets outbound from you to the router (and the internet) are working, but they're dying on the way back. If other boxes work behind that router, it would stand to reason perhaps the router isn't sending the packets back to you in response to the ones you sent out (and it does seem it did set up a session with the web server for you, or you wouldn't get a "waiting for" and we know it resolves and you can tracert it)...

The larger the L1 cache, the fewer times the CPU has to hit the (much) slower L2 cache as it can store more data in it's faster L1 cache. The larger the L2 cache, the fewer times the CPU has to hit the (FAR SLOWER) system RAM. Nanoseconds might not seem like a lot in human terms, but 8-10ns for RAM access is an eternity compared to 1-3ns hit on L1 (and even 3-5ns hit on L2). The CPU will (theoretically) not have to restart/refire instructions as frequently, and if you have a lot of cache and run a lot of data (or multiple apps with decent data sizes), you can cut down on cache-thrashing (which is probably worse than a cache miss, speed-wise).

I believe Microsoft even announced recently that most of the Vista crashes uploaded by WER were caused by Nvidia drivers, and that particular one you have was one of the worst of that lot...

This has been discussed here (and elsewhere) many times, but I won't harp on you for not searching this time because I'm in a good mood . Yes, retail, VLK, and OEM (and there's even differences in OEM versions, royalty OEM vs retail OEM) are all different and take different product keys.

You can try a repair install, which will go over and reinstall system files (like this one) and hopefully resolve that.

I'd say you can probably expect builds of Win7 sometime in 2010 (maybe late 2010), just like Vista (November 2006). Since Vista will be mainstream support until 2011, I'll consider Win7 in 2011 or 2012 unless I feel the need for a new high-end machine right around then .

Not true. Research AWE and PAE. XP (and all 32bit Windows desktop OSes, 2000/XP/Vista) are limited to 4GB of RAM, but 2GB (or 3GB with /3GB boot.ini switch) of virtual address space per process (1GB for kernel if /3GB is used).Windows 2008 x86 can utilize 32 or 64GB of RAM in certain scenarios (if the apps support AWE), and x64 much more than that.

Use runas?

Remember one thing - interviews (if they're being done well) should be about determining how you think, troubleshoot, sort logic, etc., not on what you know. You may not have the experience or the exposure to the technologies to know the answers, and interviewers do know this and as such a lot of interview questions are not looking to see if you know the answer, but how you sort through getting to one. Don't be afraid to be wrong, but try not to say "I don't know" and leave it at that - always try to come up with an answer (this should denote some mental horsepower), and if you get the question right or close to it, that's just a bonus.

The connection limit is imposed at the network layer, not the application. If you get 10 open active connections simultaneously, you will fail to host new connections. It should not be a regular issue (most web servers don't have persisitent open connections unless the client and server require it and send keep-alive headers), but it CAN happen on an XP box hosting any web server. One thing you would benefit from on Server vs XP is the longer quantum for running tasks before context swtching, better background-task memory management (for services like IIS or apache), and no limit to inbound connections (CAL limits, but not limited at the TCP stack). You could of course patch tcpip.sys to bypass this limitation, but of course that would violate the EULA and such.

Honestly, if admin accounts always work and user accounts always do not, I'd suggest running Process Monitor when trying to do certain things that always fail - if an admin can do it but a regular (limited) user cannot, it's a permissions issue (and procmon should show you perhaps access denied errors).