[4018] in linux-scsi channel archive
Re: pb. with Buslogic bt-542 card ??
daemon@ATHENA.MIT.EDU (Leonard N. Zubkoff)
Wed May 27 01:26:37 1998
Date: Tue, 26 May 1998 22:24:38 -0700
From: "Leonard N. Zubkoff" <lnz@dandelion.com>
To: doherty@thonon.inra.fr
CC: linux-scsi@vger.rutgers.edu
In-reply-to: <13674.50280.458626.65715@evian.thonon.inra.fr> (message from
Andre Doherty on Tue, 26 May 1998 17:12:57 +0200 (MET DST))
From: Andre Doherty <doherty@thonon.inra.fr>
Date: Tue, 26 May 1998 17:12:57 +0200 (MET DST)
Looks like i have a big problem to solve. I have a scsi card
Buslogic BT-542 and i am encountering severe problems with it.
On this card, there are three devices plugged: two hard disks and a
cdrom R/W Ricoh 2600S. All those things on a 486DX/2 66.
In order to try my cdrom player, i have tried to dump an audio track
from the cdrom to the hardrive. I have tried cdda2wav.
So, really self confident, i have typed :
nice -20 cdda2wav -c 2 -s -b 16 -r 44100 -t 1 -D /dev/sgc \
-I generic_scsi -v out.wav
(i have also tried readcdda, but it s the same)
All this to dump track 1 on the disk. Dump goes ok, but the resulting
file is unusable, the sound is shaky as if there was a I/O problem.
Effectively, hdparm -Tt /dev/sdb gives me:
bash-2.01# hdparm -Tt /dev/sdb
/dev/sdb:
Timing buffer-cache reads: 32 MB in 4.31 seconds = 7.42 MB/sec
Timing buffered disk reads: 16 MB in 11.19 seconds = 1.43 MB/sec
Which isn't really much !
I suspect there may be a problem with the card. But is there a way to
solve that ? Are there any tools i could use to monitor and why not
a tool to optimize the transfert rate of it ?
OR maybe it s something completely different, i am really lost.
Any idea ?
Synchronous Negotiation: Disabled, Wide Negotiation: Disabled
Target 0: Queue Depth 3, Asynchronous
Target 1: Queue Depth 3, Asynchronous
Target 5: Queue Depth 3, Asynchronous
Asynchronous, slow SCSI on an ISA bus is just not going to get you very good
performance. The BT-542B is an ancient ISA host adapter. I believe there are
jumpers to enable synchronous transfers, which should help somewhat, but
realistically you just cannot expect high performance from this configuration.
Leonard
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
