Solution for Weird Communications Issues
It would have been so easy to blame something else for the problems. I was told the problem was Windows in general. I was told it was Windows 2003 in particular. I was told it was VPC. I was told I just had a flaky system. All of these convenient answers were common solutions by people who essentially were throwing their hands up and taking guesses, but doing so with a tone that re-affirmed that their conclusions were well informed and almost certainly accurate. If I had mentioned to someone that my NIC was built-in to the mother board I’m sure someone would have questioned my choice of mobo or cited these sorts of devices as "notoriously flaky".
The problem is that technicians come up with this sort of commentary way too often. It’s easy to blame a target like Microsoft or Windows. It’s a safe bet, let’s face it, and no one can prove you wrong. It’s easy to say the problem is system-specific or bad hardware, even if no one has done any kind of hardware testing to verify the conclusion. For some people it’s preferable to save face by blaming something that can’t be proved, than to admit that they don’t have a clue what the problem is. An "I don’t know" response is OK when we ask friends or colleagues to put another set of eyes on a problem. An end-user should never be left with an "I don’t know" response, unless it’s followed with "but I’ll do what I can to find out". I’m never afraid to tell my clients "I don’t have a clue" about something because they know where we all stand, and that I’m completely honest with them. On the other hand, I rarely give that response, it’s usually "I need to do some research and I’ll let you know."
That brings us to how this sort of research is done. Flaky issues, by definition, are difficult to describe, trace, or reproduce. Doing keyword searches in Google or other search engines is an amazing way to get almost any answer you want, provided you can ask the question properly. Searches that included these keywords for my connectivity problems were largely fruitless: network, stops, pauses, dies, VPC, intermittent. It’s no wonder I couldn’t find any answers except with a detailed forum posting.
And as to a solution to finding flaky problems? One thing that comes to mind is that there are no standards for getting diagnostic data from software or hardware. My NIC was turning off and on without a single log being filed (to my knowledge). This happens all the time with Windows Update, Norton AntiVirus, email readers, browsers, Windows services, *nix background processes, etc. If all hardware and software subscribed to a single diagnostic standard for activity logging then many of these flaky problems could be resolved very quickly. Of course my comms died, look at the log, my NIC just turned off ! Maybe one of these days an RFC will be proposed for this and RFC-compliance will be perceived as a sign of quality. (Fat chance)
If you’ve read all of this then your network card is working fine. Congratulations. Look out though, those things are really completely unreli