[smartmontools-support] Power-on hours changed definition

Martin Gregorie martin at gregorie.org
Tue Nov 22 15:10:17 CET 2022


I've noticed that the method of calculating the 'power on hours',
parameter no 9, appears to have changed during the last year and am
curious about what may have changed in the way it is calculated.

Before around spring, 2022 the power-on hours parameter, which I monitor
as a warning that a drive may be approaching end of life because I've
had two or three hard drives (WD Blue and Fujitsu) fail at around 50,000
power-on hours according to smartmon.

Up the this point I noticed that power-on hours typically increase at
around 25-30 hours a week on both a server that runs 24/7 and a laptop
that's typically used most days and is quiesced overnight. This rough
synchrony is expected since the server contains both a website and a
PostgreSQL database which are fairly frequently accessed  from
thelaptop. 

However, at some point in the last year, something changed and the which
caused the power-on hours to increment at only 3-5 hours a week. I don't
thing mu use pattern has changed much, so I'm curious to know if either
you've changed the way that power-on hours are calculated or if you're
aware of changed to Linux disk handling that could have this effect.
Both computers run RedHat Fedora Linux with an XFCE desktop and both
machines get backed up and Linux packages updated on a weekly basis, so
at any time they are both running thre same software.
 
Many thanks for any light you can throw on this change.

Martin Gregorie



More information about the Smartmontools-support mailing list