Gary Roach wrote: > Thanks for the help. The real fix is as follows: > 1. The WD hard disks really are set to an 8 second head time > out. Three of my disks were beating themselves to death. This was > determined by installing smartmontools and running smartctl -a > /dev/yourdrive. Find the section of the print out that has "Power On > Hours" and "Load Cycle Count". If the "Load Cycle Count" ( the > number of head dockings) is a lot greater than the "Power On Hours" > you have a problem. > 2. Install idle3-tools and run idle3ctl -g /dev/sd?. This will > printout the idle3 timer. Run idle3ctl -h to find out the settings > and rerun idle3ctl -s <number> /dev/sd? to reset the number. Eighty > is equiv to 8 seconds 129 is equiv to 30 seconds. I set my HDDs to > 60 seconds. Turn the system off for about 10 seconds and reboot. > Fixed. This may happen for more than just WD drives. Seagate drives also have this problem. But idle3 is only available for WD drives. If you find yourself having this problem with Seagate drives then use the 'hdparm -B VALUE' with a VALUE of 254 or 255 or some such. hdparm -B 254 /dev/sda However that needs to run at boot time because it is not persistent across a power off. $ man hdparm.conf # editor /etc/hdparm.conf Bob
Attachment:
signature.asc
Description: Digital signature