[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Thu, 28 Jan 2010, Julian Hall wrote:
Gordon Henderson wrote:There is a recently published Google study of their disk failure modes, and they found that running them hot was better for them in the long-run...I know that one data robot storage company do this too with a speed/temperature controlled fan..GordonWould it be true to say that consistency of temperature is more important than actual - within reason of course? I'm thinking of the varying conditions producing more wear on all components as they constantly adjust.
It doesn't appear to be. http://labs.google.com/papers/disk_failures.pdf is the report.I've just re-read the temp. section. More disks at 30C fail than disks higher (or lower). My failling disks:
/dev/sda: WDC WD6400AAKS-00A7B0: 37°C /dev/sdb: WDC WD6400AAKS-00A7B0: 37°C Gordon
-- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/linux_adm/list-faq.html