Date: Wed, 28 Feb 2001 23:49:33 +1100 From: Zero Sum <count@shalimar.net.au> To: Niek Bergboer <niek@wit379112.student.utwente.nl>, freebsd-stable@FreeBSD.ORG Subject: Re: Adaptec AHC / Tape drive problems on 4.2-STABLE Message-ID: <01022823493300.00535@shalimar.net.au> In-Reply-To: <20010228131040.E71226@kaidaw.student.utwente.nl> References: <20010228131040.E71226@kaidaw.student.utwente.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday 28 February 2001 23:10, Niek Bergboer wrote:
>Hi all,
>
>Currently I'm having lots of trouble with the department's HP DLT-40
tapestreamer. The fileserver to
>which it is attached was formerly running FreeBSD 3.5-STABLE (from around
november 2000), and the tapest
>reamer functioned without
>any problems. Two weeks ago I upgraded to FreeBSD 4.2-STABLE (i.e. I
didn't perform a direct upgrade,
>but installed 4.2 on the first HDU leaving the second (data) HDU
untouched). The only thing that
>basically changed is the operating system: all cables and cards in the
system were untouched.
>
>Since then the tapestreamer won't function. While booting, it complains
about SCSI-timeouts (see
>attached dmesg excerpt). When issuing mt-commands, the server produces
either a nice kernel panic and re
>boots, or times out after a while.
>The machine runs the FreeBSD 4.2-STABLE from Feb 19 (compiled at Feb 20),
but similar problems existed w
>ith an earlier STABLE (about Feb 15). The SCSI-card is a Adaptec 2940A
Ultra,
>and the tapestreamer is the only device attached to it (the disks are ATA
due to budget constraints).
Nick,
I don't know if this is going to help, SCSI can be a black art
sometimes, but I had problems going from 3-5-STABLE to 4.2-STABLE with an
Adaptec 2940UW (onboard). The most significant thing was that the
frequency dropped from 20MHz to 10MHz halving transfer speed. I never got
around to analysing the problem (as other more serious issues arose) but I
co-incidently replaced the cable with a cable that was actively terminated
(not by a switch on the last device or by a resistor). When I fitted the
new cable the problem (and sundry grumblings) vanished. It might be that
4.2 is more sensitive to termination conditions, particularly if you are
manipulating wires to decrease the number of needed pins with the
"narrow-to-wide convertor plug needed". The 390U2W might be a little more
intelligent about termination.
Disclaimer: Speculation based on a single observation stretched to fit the
facts.
Zero Sum
--
count@shalimar.net.au Vescere bracis meis
http://www.hack.gr/jargon/html/D/disclaimer.html
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?01022823493300.00535>