Murphy strikes again

Interesting LKML post by Alan Cox listing all of the possible things that can cause silent data corruption when dealing with multi-petabyte (!) data sets. See also this pdf from CERN detailing the original observed problem.

Leave a Reply