On Tue, 2 Mar 2004, Erik E. Fair wrote:

> At 8:10 -0500 3/2/04, Eric Benoit wrote:
>
> >sd1: sync (50.0ns offset 15), 16-bit (40.000MB/s) transfers, tagged
> >queueing
>
> That means the drive was detected as being Single-Ended, Ultra-Wide,
> Not LVD Ultra2 Wide (that would be 80.0MB/s). Are you using the
> correct type of SCSI bus terminator for LVD? Is the drive jumpered to
> use LVD instead of SE?

I'm using an SE/LVD terminator on an SE cable, attached to the SE
connector.

> Note also that under ideal conditions, the very best you'll get is
> 30MB/s out of this drive:

I'm not expecting complete saturation, but 3M/sec with 99.5% of the time
in the kernel does NOT sound right. This would almost certainly limit the
usefulness of the system.

> Secondly, you might want to try the esiop(4) driver instead of
> siop(4); it was split out from siop to take better advantage of the
> 53c875 and higher.

I will give this a try.

Thanks

--
"It does not upset me to be 'thirteen at | eric@oddity.ca
table' unless there is only food enough  | http://www.oddity.ca/
for a dozen." - Grimod de la Reyniere    | ECSL - What conspiracy?