Just had an interesting episode with my Qube. I
noticed some performance degradation yesterday which I
figured was related to the fact that my drives were
nearly full. So I started cleaning up the drives and
but performance continued to degrade. I checked dmesg
and...

Dmesg reported this: (just a small sample #'s float a
lot)

pciide0:0:1: lost interrupt                           
        type: ata tc_bcount: 16384 tc_skip: 0         
                                   
wd1a: device timeout reading fsbn 186141632 of
186141632-186141663 (wd1 bn 18614                     
                           1632; cn 184664 tn 5 sn 5),
retrying

Smells like a hardware problem... Google brings back
lots of bad drive stories. Anyway, the problem is
isolated to wd1 so I doubt it's a cable issue. 

The interesting part is that after a reboot into
single user I was still seeing the problem and wd1 was
practically unusable. But after a power cycle I was
able to FSCK wd1 and it had no errors. Now I'm back up
and not seeing the problem, everything is working
normally and wd1 is once again streaming audio with no
problem. Spooky.. 

Anyone seen anything like this? Load is consistantly
high on my box and I haven't changed anything for
months so I'm not sure what the cause of this could be
(other than the hard drive makeing a pre death
rattle). I'll post back if I have any more info.

-Shon
http://shon.org/qube


__________________________________
Do you Yahoo!?
New Yahoo! Photos - easier uploading and sharing.
http://photos.yahoo.com/