Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

SCSI Error with mly

6 views
Skip to first unread message

Massimo Lusetti

unread,
Nov 16, 2001, 4:48:49 AM11/16/01
to

I got this errors on my 'kernel logs' mail message:
---
> mly0: physical device 0:0 sense data received
> mly0: sense key 3 asc 03 ascq 00
> mly0: info 01c04c4b csi 00000000
> mly0: physical device 0:0 sense data received
> mly0: sense key 3 asc 03 ascq 00
> mly0: info 01c04c43 csi 00000000
> mly0: physical device 0:0 sense data received
> mly0: sense key 3 asc 03 ascq 00
> mly0: info 01bd5e2f csi 00000000
> mly0: physical device 0:0 sense data received
> mly0: sense key 3 asc 03 ascq 00
> mly0: info 01bd5e2f csi 00000000
---

I use a:
---
mly0: <Mylex AcceleRAID 170> mem 0xfbffe000-0xfbffffff irq 9 at device 10.1 on
pci0
mly0: AcceleRAID 170 , 1 channel, firmware 6.00-13-00 (20010911), 32MB RAM
---
on a 4.4-STABLE box.

If i've understand correctly what's the meaning of that kernel message it
is "MEDIUM ERROR", but what really "MEDIUM ERROR" means ?! My hd is leaving
me ?! I've to get worried about that !?

I've also noted that sys/cam/scsi/scsi_all.c was updated 9 days ago since
August so do ou suggest a cvsupdate ?!

Regards,
Massimo Lusetti

To Unsubscribe: send mail to majo...@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message

Bernd Walter

unread,
Nov 16, 2001, 7:47:14 AM11/16/01
to

It would be interesting what asc 0x3 means.
But obviously your disk is failing - you should replace it.
It may also be some power supply isue.

> I've also noted that sys/cam/scsi/scsi_all.c was updated 9 days ago since
> August so do ou suggest a cvsupdate ?!

From the description Rev 1.28 should eliminate command errors, which is
different from what you see.

--
B.Walter COSMO-Project http://www.cosmo-project.de
ti...@cicely.de Usergroup in...@cosmo-project.de

Matthew Jacob

unread,
Nov 16, 2001, 12:00:34 PM11/16/01
to
>
> It would be interesting what asc 0x3 means.

blade > /etc/LGTOuscsi/ascdcode 0x3 0
ASC/ASCQ(0x03/0x00)=<Peripheral Device Write Fault>

Massimo Lusetti

unread,
Nov 16, 2001, 1:36:07 PM11/16/01
to
Quoting Matthew Jacob <mja...@feral.com>:

> >
> > It would be interesting what asc 0x3 means.
>
> blade > /etc/LGTOuscsi/ascdcode 0x3 0
> ASC/ASCQ(0x03/0x00)=<Peripheral Device Write Fault>

Where did you find this info, i would like to learn more.

Regards,
Massimo Lusetti

Matthew Jacob

unread,
Nov 16, 2001, 1:49:12 PM11/16/01
to

The tool is something I wrote for Legato years ago.

The documentation can be found in the SCSI-2 and other specs
at http://www.t10.org.

You can also find stuff somewhere down in the bowels of CAM in
sys/cam/scsi/scsi_all.?

I asked Mike why the mly driver didn't use the printout routines in CAM, and
he had the very reasonable answer of "didn't know they were there"- no
documentation means people can miss stuff.

-matt

Massimo Lusetti

unread,
Nov 16, 2001, 2:15:38 PM11/16/01
to
Quoting Matthew Jacob <mja...@feral.com>:

> You can also find stuff somewhere down in the bowels of CAM in
> sys/cam/scsi/scsi_all.?

Well, looking at sys/cam/scsi/scsi_all.c
---
/* DTL W SO */{SST(0x03, 0x00, SS_DEF,
"Peripheral device write fault") },
---

You were right, thanks for the hint.
Now i'm quite convinced my drives is leaving me :[ also if i've had 'only' five
errors in a day from it.



> I asked Mike why the mly driver didn't use the printout routines in CAM,
> and
> he had the very reasonable answer of "didn't know they were there"- no
> documentation means people can miss stuff.

Completely agree.
There's a plan to implement an interface for the mly(4) driver to control at
runtime the operation of that controller ?
I've seen Mylex porting some stuff to the Linux world...

Matthew Jacob

unread,
Nov 16, 2001, 2:17:15 PM11/16/01
to

On Fri, 16 Nov 2001, Massimo Lusetti wrote:

> Quoting Matthew Jacob <mja...@feral.com>:
>
> > You can also find stuff somewhere down in the bowels of CAM in
> > sys/cam/scsi/scsi_all.?
>
> Well, looking at sys/cam/scsi/scsi_all.c
> ---
> /* DTL W SO */{SST(0x03, 0x00, SS_DEF,
> "Peripheral device write fault") },
> ---
>
> You were right, thanks for the hint.
> Now i'm quite convinced my drives is leaving me :[ also if i've had 'only' five
> errors in a day from it.
>
> > I asked Mike why the mly driver didn't use the printout routines in CAM,
> > and
> > he had the very reasonable answer of "didn't know they were there"- no
> > documentation means people can miss stuff.
>
> Completely agree.
> There's a plan to implement an interface for the mly(4) driver to control at
> runtime the operation of that controller ?
> I've seen Mylex porting some stuff to the Linux world...
>

No clue as to this latter. Ask the author.

0 new messages