From: Christian F. <Chr...@t-...> - 2012-11-22 19:57:27
|
Karl Pielorz wrote: > Hi, > > We have a number of Intel 320 series (e.g. INTEL SSDSA2CW160G3) SSD's being > monitored by Smartmontools. Recently we noticed a number of drives have > clocked their wearout indicators down by 1%, i.e. > > " > 233 Media_Wearout_Indicator 0x0032 099 099 000 Old_age Always > - 0 > " > > Our smartd.conf says: > > " > DEVICESCAN -p -H -l error -f -m root -S on -M test > " > > We didn't get an email for this, as I'm guessing the attribute is "Old_age" > not "Pre-fail" (which the -p monitors). You never get an email for this even if -f (check for failure of usage attributes) is specified. Threshold 0 means that this attribute never fails. > The only "Pre-fail" attributes these drives list are: > > " > 170 Reserve_Block_Count 0x0033 100 100 010 Pre-fail Always > - 0 > 232 Available_Reservd_Space 0x0033 100 100 010 Pre-fail Always > - 0 > " > > Is this likely to be enough to 'catch' these drives before they fail? Media_Wearout_Indicator may be somehow connected to these attributes, I don't know. > > I just didn't want a scenario where 'Media_Wearout_Indicator' falls down to > low values, and we don't get any warnings. Add '-u -r 233! -R 233!' to the smartd.conf line to enable warning emails each time this attribute changes. See also: http://sourceforge.net/apps/trac/smartmontools/ticket/248#comment:1 Thanks, Christian |