[Fwd: Re: [Beowulf] ECC exerciser/exorciser?]

Bill Broadley bill at cse.ucdavis.edu
Mon Jan 26 15:21:23 PST 2009


I was puzzling over the best approach for testing memory and found:
  http://www.memtest86.com/tech.html#philo

In particular:
 Memory chips consist of a large array of tightly packed memory cells, one for
 each bit of data. The vast majority of the intermittent failures are a result
 of interaction between these memory cells. Often writing a memory cell can
 cause one of the adjacent cells to be written with the same data.

They discuss one algorithm, then an improved algorithm one that works even
with caches.

Seems plausible, sure a custom kernel will allow testing a larger fraction of
ram.  Seems like a good place to start for a user space test for a live
system, not sure if any of the other programs mentioned have a similar approach.

Comments?





More information about the Beowulf mailing list