>>>Also, Dmitry's theory is that the memory is corrupt. Yeah, maybe that's possible, but in my lifetime I've never encountered that too many times if ever.
>>
>>Back in the Ethernet days, we apparently had gotten a batch of bad NIC. How do I know this? Because suddenly an application that hasn't had a change done to it in over a year is getting 'Unable to access' errors and "line noise" being saved to the dbf's.
>
>As I understanding networking, those types of errors are unlikely to be the result of a hardware issue, but rather a software driver issue for that hardware.
>
>The hardware is supposed to be designed in multiple layers called a stack. Each layer is only responsible for communicating one type of information, and only to the layers directly above or below. The bottom-most layer is the copper layer (or wireless) which conveys the signals, but other ones above it are designed to be part of various protocols that are explicitly designed to transmit data in packets, and to verify the integrity of the data within the packet when sending.
>
>If, for example, there is line noise and the packet comes in with invalid information, the various protocols and layering are supposed to force it to re-send the damaged packet until it gets it corrected, and depending on which layer it was int, potentially respond back to the OS with a fatal error after so many unsuccessful retries.
>
>Since all of the new network cards were having an issue, I'd suggest it was a bad network driver for those cards, and not physically bad NIC hardware. The driver was not responding correctly to the data errors, allowing for garbage to be processed through as though it were the valid data, hence the dbf data corruption.
>
>Still, that's just a guess and I could be wrong. It's just that there are so many layers designed to keep that from happening that it would be hard to see it just being a hardware issue, and not actually a software (firmware) or driver issue.

Yup, that was the argument and the drivers were updated/replaced and I had to prove that there was nothing wrong with my software.

They were wrong. When all else is eliminated, the impossible becomes probable...or sommat like that