Re: Hard drive failures

[DrFranken <midrange@xxxxxxxxxxxx>]

Fought a similar situation with a 520 some time back. Problem was the 
disk backplane. Cheapest part in the entire setup too. CE was going nuts 
and had swapped the RAID card and many drives. Swapped out the backplane 
and problem went away.  Worth a look.

Those speaking about your drives being too old and you're living on 
borrowed time and all?  I'm not buying it. There are 10s of thousands of 
those old disks out there spinning that have been going for years and 
years and years. Remember MTBF is MEAN it does NOT mean that once you're 
past that number your drive is due to fail.

Frankie for example is running on 6718s which have been spinning for way 
over a decade and even with 45 drives fails a couple a year.

Would I configure hot spare if I could? Don't be silly, of course I 
would! Raid 6? If I could afford the performance penalty then yes that 
would be better still.  Raid 6 with hot spare? Again if you have the 
space and can handle the performance why not sleep better at night.

        - Larry "DrFranken" Bolhuis

www.frankeni.com
www.iDevCloud.com
www.iInTheCloud.com

On 2/25/2014 6:39 PM, Jeff Crosby wrote:

> All,
>
> System i 520 V7R1
> Up to date on PTFS
>
> System unit has 8 4326 drives Raid 5.
> Expansion unit has 8 4326 drives Raid 5.
>
> Since September, we've had 5 drive failures:
>
> 2013-09-23  DD003
> 2013-12-18  DD005
> 2014-02-14  DD004
> 2014-02-22  DD003
> 2014-02-25  DD001
>
> All have been in the system unit.  DD003 twice.  Note that 3 of the
> failures are in the last 11 days.
>
> The one that just happened tonight was while bringing up the system after
> replacing the internal 5755 tape drive.  To add insult to injury, the
> system is not recognizing the tape drive, so when the Raid 5 is done
> rebuilding we get to start over with the tape drive.
>
> What's causing these drives to fail?  Each drive having an isue?  Or
> something else in the system?