Re: watch your ECC memory
Umm. I didn't see the joke icon, but I'm still hoping that you're not serious about that comment.
ECC is the means to identify and correct memory errors caused by external influences. A single bit error in an ECC protected byte or word will be reported and corrected. It should not be an issue, and if it really is a bit-flip caused by a cosmic ray, is nothing to be alarmed by, the ECC has done it's job. If you get a two-bit flip, then it will be detected, but not necessarily corrected, but the system may take some action.
Some memory has a single parity bit and will detect single bit errors, but can't correct them, and may completely miss an even number of bit-flips in the protected byte or word.
Now if you continuously get errors flagged in an ECC DIMM, that probably means either that the system is on Mercury, next to a strong gamma ray source or has an alpha particle source inside the case, or that the DIMM is going bad.
Non ECC and/or parity memory is still subject to bit-flips caused by radiation, you just don't necessarily know that they've happened unless it produces noticeable data errors, or a program or the system as a whole crashes because the error occurred in a particularly important area of memory. But if that happens, you don't necessarily know that it was a memory error rather than a code error.
So don't look at ECC memory being vulnerable to these things, rejoice that it is doing what it is designed to do, protecting the integrety of your system.
Biting the hand that feeds IT
