[mythtv-users] Re: Hard Drives that Actually Work?

Greg Cope gregcope at gmail.com
Tue Dec 28 21:26:48 UTC 2004


Sorry, meant no.

Most disk failures that I have seen (100+ disks, in DC's that have
1000's) have been when the systems have been power cycled, and the
disks fail at startup or just after.

Normal policy at some places is to replace at the the 1st sign of
errors, as you can factor this work in when you want to, compared to
it failing when you do not want it to (Friday afternoon).  The cost of
the disk is actually small compared with the work required to replace
it properly (take a backup, replace disk, check it etc ... A sysadmin
has to do some work, along with a disk swapping monkey (might be one
and the same!))

I've only seen a few (less than 10, probably 5) fail whilst in use. 
So a rough guestimate would be that 95% of disks failures I have seen
are at power up.

I am supplier agnostic, with Seagate and IBM/Hitachi being a
perferance at the mo.  But have Samsung/WD drives too.  Oh - I also
have lots of backups, my preferance being a USB enclosures ~ these are
cheap and mobile.

Greg


On Wed, 29 Dec 2004 00:06:44 +1100, Hamish Moffatt <hamish at cloud.net.au> wrote:
> On Tue, Dec 28, 2004 at 10:00:39AM +0000, Greg Cope wrote:
> > On Tue, 28 Dec 2004 12:43:33 +1100, Hamish Moffatt <hamish at cloud.net.au> wrote:
> > > Does a disk which is spinning 24x7 but is mostly idle
> > > (eg my Myth installation) fail more quickly than a disk
> > > which is powered off for those other hours?
> 
> > I'd say yes, as in most datacentres I've seen more disks die at
> > spinup/start up than during normal 247 use.
> 
> Err, did you mean no? I'm confused.
> 
> Hamish
> --
> Hamish Moffatt VK3SB <hamish at debian.org> <hamish at cloud.net.au>
> 
> 
> _______________________________________________
> mythtv-users mailing list
> mythtv-users at mythtv.org
> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
> 
> 
>


More information about the mythtv-users mailing list