Andre Albsmeier wrote:

On Tue, 03-Aug-2004 at 23:31:52 -0400, [EMAIL PROTECTED] wrote:

Please cc replies directly to me, as I am not subscribed to the lists.

With some help from here, I was able to get this RAID card to see our
external DLT (QUANTUM 4000) SCSI tape drive by installing the aacp (pass
through) driver in addition to the aac driver.  camcontrol now works, as
do basic mt commands and amcheck (amanda check).

However, (amanda) dumps either hang, fail completely or fail after
transfering very little data.  On the console, I see:

(sa0:aacp1:0:4:0): READ(06). CDB8 0 0 0 20 0 0
(sa0:aacp1:0:4:0): NO SENSE ILI (length mismatch): -24576 csi:0,0,0,1

At this point the device is completely unresponsive, and the only way to
get the system to see it again is to reboot the whole server.  I tried
ordering a 3 ft cable, thinking I was pushing my luck with the 6 ft (I've
had this problem with SCSI cables in the past), but the problem persists.

The same drive (which has an active terminator) has been working fine for
years on a different box using an Intel L440GX+ MB's on-board SCSI port.

Once again, any helpful replies are greatly appreciated!


Are you sure you are running a recent fw on your DLT4k? My DLTs
used to behave badly with early fw revisions. Check out

http://www.quantum.com/am/service_support/downloads/software/dlt4000.htm

You can upgrade it by tape or use my software for updating the fw of
SCSI devices on FreeBSD.

        -Andre


This sounds like excellent advice. Note that the error messages that you are seeing are coming from the Adaptec firmware, not FreeBSD or the aac driver. Also, the aacp device and backing firmware support are really just hacks that exist to allow cdroms to be booted and drives to be flashed with new firmware. I've never heard of anyone running a tape drive in this fashion, so it will be quite interesting to see if newer firmware helps.

Scott

_______________________________________________
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to