On 2025-03-05, Stefan Monnier <monn...@iro.umontreal.ca> wrote: >> I thought the whole point of running the SMART tests was to detect >> a failing disk, so color me confused. > > If we knew how to detect a failing (as opposed to failed) disk, then > things would be easy. SMART is an attempt to provide relevant > information, but that's all. AFAIK, it's still "the best we have", but > it's not a reliable way to answer the question (I'm not even sure the > question is well-defined, to be honest).
Given the cryptic nature of its results and the fact that a positive result implies that the test itself can be a factor in precipitating the very failure it intends to preclude, I see no reason to run this program, which seems to produce more confusion than anything veritably useful. On my machine, I simply back up the data I do not care to lose. > > Stefan > >
