YIKES..  We have that problem also.  Get rid of the scsi_options line (put a
* in front of it to comment it out).  

This throttles ALL scsi buses on the systems to 40MB/sec.   I've been burned
by this many times when connecting LVD scsi (160+ MB/sec).   It seems ok
writing to a drive but when the bus talks to the drive arm I think it may
attempt to negogiate the bus speed down..newer drives then don't work at the
lower scsi bus speeds.  

//Soap box mode on//

It seems ancient/older EMC folklore..from a time when EMC connected disk at 
HVD differential?   There are whole environments of servers out there that
still have this setting?   Luckily most newer servers had fiber attached
disk and the setting does nothing..   I sometimes question also whether the
max_throttle is needed with today's fiber attached disk and large array
caches.  If fact I'm going to ask our EMC engineer if I see him today.

// rant mode off  //



Sometimes tape drives and libraries work fine with the setting but if one
checks the bus speed they will it running at 40MB/sec...a bit slow for
today's drives where one drive can do 50MB/sec +..

To check all this please try/take a look at John DiMarco's scsiinfo.. as of
version 4.7, it did report qus (SUN LVD) info..  The last link I have is:
http://www.cs.toronto.edu/~jdd .    Solaris prtconf can also you give scsi
bus speeds..albeit in hex.


If taking the setting out the issue gets worst than it means there is some
issue with the LVD cabling/termination that can't handle the higher speeds.


Good luck.


Robert Maiello
Pioneer Data Systems



-----Original Message-----
From: Bolay, Vincent [mailto:vincent.bolay AT medtronic DOT com] 
Sent: Friday, July 22, 2005 4:47 AM
To: Legato NetWorker discussion; Maiello, Robert
Subject: RE: [Networker] Tapes marked full after a few gigs

This is what I have in /etc/system :

set sd:sd_max_throttle=20
set scsi_options=0x7F8
set sd:sd_io_time=0x78

These settings are needed by the EMC box. I doubt that the sd settings may
have an influence for st.
scsi_options=0x7F8 means everything enabled.

Vincent

-----Original Message-----
From: Legato NetWorker discussion
[mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU]On Behalf Of Robert Maiello
Sent: Thursday, July 21, 2005 11:23 PM
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: Re: [Networker] Tapes marked full after a few gigs


With older SDLT320 drives I had a similar problem; although new media
faired a little better.  For SDLT320 there was a calibration track
written to the media when it is first written.  If a bad drive writes
this track other drives have a problem reading it.  Based on tapes I
send to Quantum/ATL they replaced certain drives.   This doesn't
sound like your problem though.. your saying new media on either drive
will produce the write error.

It being LVD scsi attached, I'd definitely check/swap out cables and
terminators.  Check /etc/system and make sure you have no scsi-options
line throttling the bus.

Robert Maiello
Pioneer Data Systems


On Wed, 20 Jul 2005 11:37:10 +0200, Bolay, Vincent
<vincent.bolay AT MEDTRONIC DOT COM> wrote:

>Hello,
>
>This is my config:
>
>SunOS lau-backup2-b 5.9 Generic_117171-05 sun4u sparc SUNW,Sun-Fire-280R
>Patch: 113277-24 SunOS 5.9: st, sd and ssd Patch
>Sun StorEdge(TM) Enterprise Backup 7.2.Build.172 Network Edition/35
>Storagetek Library L80 with Quantum SDLT320 drives firmware rev 80, SCSI
LVD
>
>I often get the following messages:
>
>/nsr/logs/daemon.log:
>
>07/20/05 01:32:41 nsrd: media notice: sdlt320 tape DEX066 on /dev/rmt/0ubn
is full


>07/20/05 01:32:41 nsrd: media notice: sdlt320 tape DEX066 used 28 GB of 160
GB capacity
>07/20/05 01:32:59 nsrd: media notice: Volume "DEX066" on device
"/dev/rmt/0ubn": Block size is 32768 bytes not 131072 bytes. Verify the
device configuration. Tape positioning by record is disabled.
>
>/var/adm/messages:
>
>Jul 20 01:32:41 lau-backup2-b scsi: [ID 107833 kern.warning] WARNING:
/pci@8,700000/pci@1/scsi@2/st@1,0 (st50):
>Jul 20 01:32:41 lau-backup2-b   Error for Command: write
Error Level: Fatal
>Jul 20 01:32:41 lau-backup2-b scsi: [ID 107833 kern.notice]     Requested
Block: 9579                      Error Block: 9579
>Jul 20 01:32:41 lau-backup2-b scsi: [ID 107833 kern.notice]     Vendor:
QUANTUM                           Serial Number:  P    i
>Jul 20 01:32:41 lau-backup2-b scsi: [ID 107833 kern.notice]     Sense Key:
Media Error
>Jul 20 01:32:41 lau-backup2-b scsi: [ID 107833 kern.notice]     ASC: 0xc
(write error), ASCQ: 0x0, FRU: 0x0
>
>It can happen on any drive and with new or used tapes. The tapes are
labeled always the same way.
>I know that this is a known issue, but nobody (Sun, Sunmamagers, Google)
has been able to provide a solution yet.
>
>TIA for your replies
>
>Vincent Bolay
>Systems Engineer
>Medtronic Europe S�rl
>Route du Molliau 31
>Case postale
>CH-1131 Tolochenaz
>Switzerland
>Phone: (+41 -21) 802 75 34
>Fax: (+41 -21) 802 79 11
>E-mail: vincent.bolay AT medtronic DOT com
>
>--
>Note: To sign off this list, send a "signoff networker" command via email
>to listserv AT listserv.temple DOT edu or visit the list's Web site at
>http://listserv.temple.edu/archives/networker.html where you can
>also view and post messages to the list. Questions regarding this list
>should be sent to stan AT temple DOT edu
>=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=