Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 30 Sep 1999 15:25:27 +0800 (SGT)
From:      Gregory Hosler <gregory.hosler@eno.ericsson.se>
To:        Michael Meissner <meissner@cygnus.com>
Cc:        aic7xxx@FreeBSD.ORG, gregory.hosler@eno.ericsson.se (Gregory Hosler), noe@monet.prism.uvsq.fr
Subject:   Re: fast drives.
Message-ID:  <XFMail.990930152527.gregory.hosler@eno.ericsson.se>
In-Reply-To: <19990929093341.C4929@elmo.cygnus.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Michael, your setup is not all that different from mine.

On 29-Sep-99 Michael Meissner wrote:

> Hmmm, I've had no problems with a pair of 9gig IBM 10,000rpm DRVS09V disks (I
> think these are 9zx drives) in my x86, and my alpha just arrived with another
> 9gig 10,000rpm disks.  In both systems, I use the default 2.2.12 drivers. 
> The
> x86 has an Adaptec 7896 built into the motherboard (Intel L440GX+) which has
> two channels, and the two IBM disks are hooked up to the first channel as LVD
> drives, and I hook up 3 older disks (2 Seagate Baracuddas, 1 Quantum Viking)
> to
> the other other channel in SE mode.  The alpha has an Adpatec 7881 based
> controller in it.
> 
> Off hand, it certainly sounds like a termination and/or cabling issue.  Note,
> I
> believe most LVD capable disks need external active termination.

In as much as I tend to be more SCSI competent than most of my peers, I
frankly admit I feel very incompetent as regards to some scsi issues.

My 2 discs:

I have the following:

         (scsi0:0:0:0) Synchronous at 40.0 Mbyte/sec, offset 15.
           Vendor: IBM       Model: DDRS-39130D       Rev: DC1B
           Type:   Direct-Access                      ANSI SCSI revision: 02
         Detected scsi disk sda at scsi0, channel 0, id 0, lun 0
 
         (scsi0:0:5:0) Synchronous at 40.0 Mbyte/sec, offset 15.
           Vendor: SEAGATE   Model: ST19101W          Rev: 0011
           Type:   Direct-Access                      ANSI SCSI revision: 02
         Detected scsi disk sdb at scsi0, channel 0, id 5, lun 0

The interface is an motherboard AIC-7896.

The Seagate is, to the best of my knowledge, fast/wide only. Question:
how do I tell if it is "SE" ? (the adaptec reports it as "SE". Ignorant
question time. What's "SE" ?). I know that the Seagate is not LVD.

The IBM is jumperable LVD or SE. To be honest, I am not sure the present
setting. I have tried both in an attempt to get it to register at BIOS
probe setting at 80.0 MB/s, but to no avail. I have seen other posters
using this drive, so there must be something I am likely doing wrong.

more ignorant questions:

1) As I am learning, I am now (only now) guessing that it might not be a
   good idea to combine LVD and SE on the same scsi bus.
   is this true ?

2) are there any special cabling / termination issues with respect to LVD
   that are different from SE scsi ?

3) How do I get my AIC to recognoze the IBM at 80 Mbs transfer rates ?
   (do I need to move the Seagate to a different scsi bus ?)

Cableing:

The scsi's are in an external box. The cabling in the box is blue and flat
but not the typical "flat ribbon" cable. supposedly it is specifically for
the faster disks, but I'm not sure what it should really be, and I don't
know (for sure) what it really is, except that it is not the standard
68 wire wide flat ribbon cable. The external box is terminated with an
"active terminator" and teh other outlet of the box is connected to the
pc box by a heavy duty 100 ohm 68 pin cable. The pc box to motherboard is a
68 pin "twisted pair" cable. (lots of loose wires, nothing at all like a
ribbon cable).

The setup works much of the time, especially under no load (the
non-interesting and trivial case, unfortunately). Under load, it fails
w/ vaarious messages, usually starting w/ complaints such as the
following:

        No active SCB for reconnecting target

or some time out error.

My 1st step was to disable tagged queueing (at the driver compile level).
This significantly reduced the occurance of the problems, but not
eliminated. The next step was to reduce the transfer speed in the
Adaptec BIOS. I went from 40.0 to 32.0 - I am still getting time out
errors (though less frequently).

It occurs to me that by disabling tagged queueing, and reducing thruput
to 20mb, or less, I have totally defeated the purpose of buying a 80.0
mb drive (which at best I can only seem to get seen as 40.0 mb).

I am really discouraged, and at the point of seriously contemplating
switching adapter cards, the only thing is that the AIC is on the mother
board, and is one of the reasons I bought this particular mb.

-Greg




To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.990930152527.gregory.hosler>