I have a SS2 running NetBSD 1.5.2. Recently the system disk at ID 0 seems
to have grown a media fault:

Sep  6 09:00:26 picarefy /netbsd: sd0(esp0:0:0):  Check Condition on CDB: 0x08 02 93 c0 10 00
Sep  6 09:00:26 picarefy /netbsd:     SENSE KEY:  Media Error
Sep  6 09:00:26 picarefy /netbsd:    INFO FIELD:  168903
Sep  6 09:00:26 picarefy /netbsd:      ASC/ASCQ:  Unrecovered Read Error
Sep  6 09:00:26 picarefy /netbsd:      FRU CODE:  0xd0
Sep  6 09:00:26 picarefy /netbsd:          SKSV:  Actual Retry Count: 32
Sep  6 09:00:26 picarefy /netbsd:

   This error can be provoked by attempting to do anything with /dev/ttyp*
and /dev/ttyr[0-3], even ls them. My first thought was that the bad block
was in /dev, except that errors only occur on these entries, and fsdb says
that /dev occupies blocks 102504 and 102512. Since these are device nodes,
everything about them is packed into the dir entry, right? And the boot
partition is at the beginning of the disk, so I don't see any way to
correlate 168903 (that is the block number where the error occured, right?)
with 102504 or 102512.

   What is it trying to tell me? Other than that I should replace it soon.
Thank you!

   --James B.