On Mon, 24 Feb 2003, Bruce Allen wrote:
>> > >As far as the units that Fujitsu is using, it looks like it might be in
>> > >SECONDS. 89955871 seconds = 24987 hours = 1041 days which is about three
>> > >years. Could you try and confirm this by running smartctl -a, waiting a
>> > >five minutes, and running it again, and then seeing if the raw value
>> > >increments by ~300?
>> > Seem to be seconds, yes. It has increased ~230000 over the weekend, which
>> > is ~63 hours.
>> Do you want to code a:
>> -v 009,seconds
>> command-line option, and update the docs? Or do you want me to do this
>> (and the you test it).
>I had an unexpected 15 minutes free this morning, so I added a
> -v 9,seconds
>option to smartctl, and a corresponding Directive to smartd.conf. Could
>you please download it from CVS and give it a try?
Sure, since you beat me to coding it :)
./smartctl -v 9,seconds -a /dev/hda
9 Power_On_Seconds 0x0012 001 001 020 Old_age FAILING_NOW 25081h+15m+27s
... 25081h+15m+27s isn't a very nice number, but anyway, it's parsable.
To be consistent, without the -v 9,seconds, shouldn't we show
power_on_hours with a "h" suffix? But then again, calling it raw value
after our massaging and unit conversion is a bit weird :)
Or should we just change the attribute name (ie Power_On_Hours to
Power_On_Seconds) when the -v option is invoked, and do no harm to the
raw value? That seems to me to be the cleanest way... the h and m suffixes
stand out among the other values.
./smartctl -v 9,minutes -a /dev/hda
9 Power_On_Minutes 0x0012 001 001 020 Old_age FAILING_NOW 1504878h+07m
./smartctl -a /dev/hda
9 Power_On_Hours 0x0012 001 001 020 Old_age FAILING_NOW 90292636
Erik I. Bolsø | email: <knan at mo.himolde.no>
The UNIX philosophy basically involves giving you enough rope to
hang yourself. And then a couple of feet more, just to be sure.