This message is from the T13 list server.

Just in case you have never seen the SATA problems that I have been
complaining about or the Jeff Wolford has been discussing here, then
I have a test log file for you to view. This is for a 96 hour test
run on a SATA drive. I have 'x'ed out some of the drive ID data but
you can see the system is an ICH5. This log file shows 31 errors in
96 hours - this is REALLY REALLY VERY VERY BAD!!! The same test on a
PATA drive would have no errors - probably not even a single UDMA
ICRC error.

The log file is at < http://ata-atapi.com/ctlog1.txt >

You can see that some of the errors were caused by the drive
"reseting itself" - in the middle of a command the drive presents the
reset signature data. In other cases the drive's status changes to
strange values like FFH. The two data compare errors in this log may
be actual problems with this model drive, HOWEVER, I have seen data
compare errors on SATA drives that appear to be corruption of the
command parameters by the SATA interface when the command is sent to
the drive - in this case the drive thinks is has received a valid
command but the command parameters (like the LBA value) as seen by
the drive are incorrect.

[And I must add this: Back in the days before "toy computers",
problems like this would cause a "stop ship" and a massive
engineering effort to find and fix the problems. I still wonder how
these kinds of problems are allowed in products you can find in your
local computer store?]

Hale



*** Hale Landis *** www.ata-atapi.com ***



Reply via email to