hard disc temperature

[LudwigVan]

Hi there, I am asking for your experience.

I now have my B|2 for 4 years with its original 1T hard disc. I recently had a sverfailure in the office which was caused by the hard discs running over 40 °C. My B|2 hard disc allway runs at about 50 °C. So now I am a bit worried. Any experiences?

LudwigVan

[Ubi]

a server failure is not 'caused' by a disk runnign at 40 degC if it has any build quality in it. 50degC is on the edge but is generally within spec for a disk. But any disk can fail, no matter what temperature it is running. So you should keep a backup.

[Cheeseboy]

Hi,

I guess you have a WD Green in your B2. According to WD's spec sheets the operating temperature is 0-60 degrees Celsius.
Running at a higher temperature might shorten the life of the disk, but on the other hand it is like a light-bulb: it will fail sooner or later.
It might fail the day after you bought it, or after 5 years - you never know. So as Ubi said, keep a backup.

There are also tools that might give you a warning before it completely blows, giving you time to replace it before it happens. It saved me once.
I recommend you install the package smartmontools. Then you can always get the S.M.A.R.T info from your drive like this:

Code: Select all

$ sudo smartctl -a /dev/sda

It also comes with a nice feature called smartd, which you can configure to monitor your disk(s) automatically, and tell you if something starts looking fishy. You configure it with something like this in /etc/smartd.conf:

Code: Select all

# First (primary) ATA/IDE hard disk.  Monitor all attributes, enable
# automatic online data collection, automatic Attribute autosave, and
# start a short self-test every day between 2-3am, and a long self test
# Sundays between 3-4am.
/dev/sda -a -R 194 -o on -S on -s (S/../.././02|L/../../1/03) -m admin

The "-m" bit at the end means it will send an email to the admin user if something unexpected is detected (that's if you have a working mail system on your server). It will also print warnings to the syslog, so if you use something like logcheck, or manually check the logs regularly (as you should), you will see the warnings anyway.

None of this is a guarantee though. You really need to have a backup if you have data you don't want to lose.
The original disk in my B2 (I bought it the same week it was released) is still ticking along with 40000+ Power_On_Hours.
Remember all the scares about the Load_Cycle_Count on WD Greens? I have an even older WD Green in my media-player, which spins up and down constantly. I never bothered to fix it on that one, because there is nothing important on it. Still working perfectly though, after daily use since 2007.
On the other hand the disk in the B3 I bought a year or two after the B2 has needed replacing. (So again - you never know).

Cheers,

Cheeseboy