Michael Kjörling: > On 8 Oct 2024 11:29 -0400, from d...@randomstring.org (Dan Ritter): >> >> This looks like a drive which is old and starting to wear out >> but is not there yet. The raw read error rate is starting to >> creep up but isn't at a threshold. > > I agree. The almost 62000 hours is well over 7 years of run time, and > based on the start/stop count and power cycle count it's been running > basically continuously for that time (which is generally good for > longevity, as long as it's not subjected to excessive heat).
It is exactly that. It has been running mostly uninterrupted in my
basement. Max temp from the past 12 months (as far as I can tell by
looking at aggregated data in munin) is 36°C. That should be fairly
ideal.
> Also note that some disks actually lie in SMART data. I don't know if
> yours does, but I would definitely question a value of 0 for failed
> (current pending and offline uncorrectable) _and_ reallocated sectors
> for a disk that's reporting I/O errors, for example. _At least_ one of
> those should be >0 for a truthful storage device in that situation.
That is exactly what was confusing me here.
> What I would not do at this point is subject it to more physical
> stress than unavoidable. Unless you absolutely must, do not physically
> unplug or remove that disk before the RAID array has resilvered onto
> the new disk. It's currently providing value being a second source of
> truth about what's stored; you don't want to remove it and then find
> during the resilver that the other current disk has a problem.
Helpful advice, thanks. Unfortunately, I cannot hotplug into this
system. Thinking of this, the errors came shortly after a (long overdue)
reboot, so it will have to survive at least another shutdown to provide
some redundancy.
Just for completeness, the long self-check did not report any issues and
the SMART values also stayed the same. Nothing to see here, move along.
^^
The new disk is already sitting on my desk.
J.
--
I cannot comprehend the idea of chemical and biological weapons.
[Agree] [Disagree]
<http://archive.slowlydownward.com/NODATA/data_enter2.html>
signature.asc
Description: PGP signature
