[colug-432] USB disks and Load Cycle Count

[Bill Schwanitz]

make sure the drive shows up as in the database ( smartctl -a, it's towards the top of the output ). if its not, the only way to fix that I know of is to compile fresh from the main repo ( the drives are stored in a .h file somewhere )

we had that issue with some newer drives ( crucial m4 ) and that upgrade made the values actually make sense. 

short version of why is that not all drive manufacturers use the same id/property mapping. not saying this is your problem but it could be


Bill Schwanitz

If A is a success in life, then A equals x plus y plus z. Work is x; y is play; and z is keeping your mouth shut. - Albert Einstein.

On Apr 3, 2013, at 3:28 PM, Rob <res at colnet.cmhnet.org> wrote:

> I have some 2.5" USB portable hard drives that I would like to press
> into service as the main drive for several Pogoplug servers I have set
> up locally.  These drives work fine, but when I run smartctl on them,
> I'm seeing very high and rapidly increasing values for the SMART
> attribute Load_Cycle_Count.  (Rapidly increasing means several thousand
> instances per day.)  Load_Cycle_Count is an indication of how many times
> the heads have been parked onto and then moved off their landing zone.
> It seems the firmware in a lot of pocket drives automatically park the
> heads if there has not been any disk activity for 5 seconds or so.  One
> disk which had been pressed into service in January was showing over
> 250K head loads.
> 
> According to Wikipedia (the gold standard for unimpeachable data), a
> typical lifetime rating for Load_Cycle_Count is 300-600K and when used
> as I am using these disks many drives exceed their design limit within
> the first year.  This self-parking "feature" may be fine when the drive
> is carried around in a briefcase and only sporadically connected, but
> it is decidedly detrimental in a lightly loaded server application where
> there might be several disk accesses a minute.
> 
> My WD and Toshiba 2.5" drives seem to exhibit this behavior.  My
> Seagates either don't, or don't report this statistic.  I suspect
> the 3.5" drives are not so programmed, since it would be more typical
> to use them in a fixed 24/7 configuration rather than as a portable
> drive.
> 
> Has anyone else encountered this with drives you have used, and if
> so, what did you do about it?  Is there a way to instruct the drive
> to not park its heads automatically based on time in a quiescent state?
> 
> Thanks,
> Rob
> _______________________________________________
> colug-432 mailing list
> colug-432 at colug.net
> http://lists.colug.net/mailman/listinfo/colug-432
>