Re: Problems w/ Sil3124 + Port Multiplier

[Tejun Heo <htejun@gmail.com>]

Lars Michael Jogbäck wrote:
> Tejun Heo wrote:
>> I was hoping to see some error logs but no.  Hardware_ECC_Recovered
>> count seems high (385707184) but I dunno whether the value is normal or
>> not.  Different manufacturers use different norms in counting them.  If
>> you have other disks of the same model, you can compare the values and
>> see whether if it's unusually high.
>>   
> I think that is normal for that kind of disk. This is another disk but
> the same model (this one is attached to a 3ware 9500-controller)
> and it shows the same.

I see.

> May  1 23:53:36 cleopatra kernel: ata5.00: tag 0 cmd 0xea Emask 0x4 stat
> 0x40 err 0x0 (timeout)
> 
> and then it continues with approx 1/hour of the above. So it seems
> something is strange in the interface between the drive and the
> computer. I can swap the drive (it's in an raid5-array) to another drive
> of the same model if helps in any way; but I suspect that it will show
> the same result.

The fact that the first slot is always the one causing problem bothers
me.  Can you try to connect the disk to a different controller and use
different power connector?  Let's see to which the problem is attached -
disk, port or something else.

-- 
tejun