Re: Lots of con-current I/O = resets SATA link? (2.6.25.10)

[Gerhard Wiesinger]

On Wed, 9 Jul 2008, Gerhard Wiesinger wrote:

> On Tue, 8 Jul 2008, Justin Piszcz wrote:
>
> > > > > > Hi Gerhard,
> > > >
> > > > Please run:
> > > >
> > > > smartctl -t short /dev/sdc
> > > > sleep 300
> > > > smartctl -t long /dev/sdc
> > > >
> > > > Wait 2-3 hours or more and:
> > > >
> > > > smartctl -a /dev/sdc
> > >
> > >
> > > I'm changing the disk one more time ...
> >
> > Once you replace the disk please re-run the tests shown above and show us 
> > the output please.
> >
> > How many SAMSUNG disks have you gone through, do they really fail that 
> > often?
>
> Hello Justin!
>
> Setup Power suppply:
> sda-sdd are on the same power supply cable, sde, sdf are on a second one.
>
> I changed the disk to a new SAMSUNG disk and the first raid resync did NOT 
> work and I had 1 reallocated sector ...
>
> Also the RMA guy at our very good computer store (http://www.ditech.at/) were 
> helpfull in changing the disks and he hates the SAMSUNG disks because of high 
> RMA rates ...
>
> But:
> smart checks were ok (details see below). The smart checks are shown as 
> aborted, but before running the next tests they were at "Completed without 
> error" (strange).
>
> Now sdc has it's own power cable and I'm running a raid resync now.
>
> Power supply is:
> ATX 385 Watt, ENERMAX PRO82+
>
> Load on power supply measured on 220V side with a true RMS Wattmeter:
> Power off: 6,5W
> Max seen at power up: 196W
> BIOS: 114,1W
> Linux idle: 92,0W
> 100% Disk I/O, 100% CPU load: 149,6W
> => Everything fine and as expected.
>
> I'll keep you up to date ...

Hello!

It looks like it works now without any problems:
*) Serveral RAID checks (once a day)
*) smartctl checks were fine
*) Only one reallocated sector
*) Writing the disk full with random data

So the possibilities for the fault are:
*) SATA power connector (most likely I would guess)
*) 4th disk is now ok
*) Other 220V power cable connector through the Wattmeter

Ciao,
Gerhard

--
http://www.wiesinger.com/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/