[Dirvish] Failing swap cage

Keith Lofstrom keithl at kl-ic.com
Sat Feb 12 07:39:43 PST 2005


In the last few days, I started seeing this error in the log files
of my backup server:

Feb 10 04:13:16 hdg: dma_intr: status=0x51 { DriveReady SeekComplete Error } 
Feb 10 04:13:16 hdg: dma_intr: error=0x84 { DriveStatusError BadCRC } 

hdg is the backup drive.   I got about 200 of these a night, during
backups of course.

I swap my backup drives every day, though I stopped when I saw this 
error, so I would not risk my other backup drives until I tracked
this down.  Fortunately, there was no apparent damage to the data
on the drive.

The error stopped when I replaced the swap cage hardware.  So, either
the electrical contact pins in the cage have worn out, the fan on the
swap has stopped, or the IDE connector in the back of the cage was
partly loose.  This kind of thing can happen with poor electrical
contact on the IDE connector - the signals become intermittent. 

And that can happen when the cage is being slammed with a new drive
removal/insertion every day - connectors can lose their gold plating,
or IDE cables can work their way loose.  There is a lot of weight
hanging on the IDE plug, and the plug and socket get flexed every
time the cage knocked about.

When I get time, I will narrow the cause down further, but for now,
if you see this kind of problem, check your IDE cable connections.

Keith

-- 
Keith Lofstrom          keithl at keithl.com         Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs


More information about the Dirvish mailing list