SCSI command timeout

SCSI command timeout

Post by David W. Shar » Fri, 17 Feb 1995 08:52:42



Using  tar xf /dev/rst11 I get the following error:


                Disconnected command timeout for Target 4.0

                SCSI transport failed:reason 'timeout': giving up

The tar device is a 4mm Exabyte 4200 and it is attached to the second
SCSI Sbus card and is the only device attached to that card.

What gives?

The Solaris 2.3, st.conf file is setup with:

   tape-config-list="EXABYTE EXB-4200","EXABYTE 4200 4mm DAT",EXBT-42c";
EXBT-42C=1,0x34,512,0x0039,1,0x13,0;

------------------------------------
David W. Sharpe                                

Consistency Point Technologies, Inc.
(408) 263-7479
(408) 263-0925 (FAX)


 
 
 

SCSI command timeout

Post by Simon Lockha » Sat, 18 Feb 1995 19:52:42




>Using  tar xf /dev/rst11 I get the following error:

>            Disconnected command timeout for Target 4.0

>            SCSI transport failed:reason 'timeout': giving up

I am also getting a similar problem. Our backups are run automatically
overnight, using a combination of 'cpio' and 'dd'. At random, the backups
would fail, leaving messages such as:


   Disconnected command timeout for Target 4.0

   SCSI transport failed: reason 'timeout'
: giving up
   polled command timeout
esp:               State=MSG_OUT_DONE Last State=MSG_OUT
esp:               Latched stat=0x96<IPND,XZERO,MSG,CD> intr=0x18<BUS,FCMP> fifo 0x20
esp:               last msg out: IDENTIFY; last msg in: COMMAND COMPLETE
esp:               DMA csr=0xa6240210<EN,INTEN>
esp:               addr=fc000dea dmacnt=8000 last=fc000de8 last_cnt=5
esp:               Cmd dump for Target 4 Lun 0:
esp:               cdblen=6, cdb=[ 0x10 0x0 0x0 0x0 0x1 0x0 ]
esp:               pkt_state=0x3<SEL,ARB> pkt_flags=0x0 pkt_statistics=0x0
esp:               cmd_flags=0x10022 cmd_timeout=540

The tape drive (Sun 5Gb 4mm DAT), would then be left with one of the LEDs
flashing.

At first, I thought this might be related to the way we are doing the
backups, but this morning I tried doing a 'ufsdump' to compare backup
speeds, and the same thing happened, with the machine also locking up when
the errors started. Power-cycling the tape drive got the machine working
again.

The tape drive is on the same SCSI bus as 2x1.05Gb and 3x2.1Gb drives and a
CDROM drive. The machine is a SS20.

Has anyone else come across these problems? Has anyone else managed to fix
the problem?

Thanks,

Simon
--


Technical Services      |
British Gas Service     |

 
 
 

SCSI command timeout

Post by Doug Gwy » Sat, 18 Feb 1995 21:47:16




>>Using  tar xf /dev/rst11 I get the following error:

>>                Disconnected command timeout for Target 4.0
>I am also getting a similar problem. Our backups are run automatically
>overnight, using a combination of 'cpio' and 'dd'. At random, the backups
>would fail, leaving messages such as:

>   Disconnected command timeout for Target 4.0

This sounds similar to the bug # 1122712 which was fixed for Solaris 2.2
in patch 101228-03.  Have you looked for a similar patch for whatever
release you're running?
 
 
 

SCSI command timeout

Post by Dr.Ivan S. Bisho » Wed, 22 Feb 1995 17:21:42


It may be a red herring, but I had similar problems when using tar and
ufsdump on 2.4X86 and 2.1X86. I found that by limiting the blocksize to
64K stopped the problems.
--
Ivan S. Bishop

 
 
 

SCSI command timeout

Post by Darren Morg » Sat, 04 Mar 1995 00:50:04


I also have had

        SCSI transport failed: reason 'timeout'

with the SCSI activity light lit on the tape drive. In my case,
the problem was a faulty SCSI cable to the tape drive.

Darren

 
 
 

SCSI command timeout

Post by Tikkana Akura » Sat, 04 Mar 1995 00:22:22



: : I am also getting a similar problem. Our backups are run automatically
: : overnight, using a combination of 'cpio' and 'dd'. At random, the backups
: : would fail, leaving messages such as:


: :    Disconnected command timeout for Target 4.0

: :    SCSI transport failed: reason 'timeout'

: : The tape drive (Sun 5Gb 4mm DAT), would then be left with one of the LEDs
: : flashing.

: I am having exactly the same problem.  It used to work fine in the past,
: but now it fails _always_.  This means that we are currently unable to
: make backups at all!!!


:       Disconnected command timeout for Target 4.0
: WARNING:

:       SCSI transport failed: reason 'timeout': giving up

: The machine in question is SPARC classic with a DAT drive.  The following
: patches are applied:

:  101316-01 101317-04 101317-09 101317-10 101318-36 101318-59 101327-02
:  101327-08 101329-09 101331-03 101331-04 101331-05 101344-05 101344-11
:  101347-01 101347-02 101362-09 101378-10 101415-02 101533-03 101533-05
:  101615-01 101615-02 101739-04 101889-01 101889-03 102034-01 102167-01

: Michael

I use ufsdump for backups, also automated to run at night. I never got the
'timeout' warning at all, but ofthen I get the following message:

WARNING:

        Soft error rate (retries = 1) during writing was too high
        Periodic head cleaning required and/or replace tape cartridge

This warning comes even after the head was cleaned with a cleaning cartridge the
before inserting the backup tape!

--
**************************************************************************
Tikkana Akurati, M.S.                     Ph-1: (215) 898 - 9416
System Administrator/Programmer Analyst   Ph-2: (215) 898 - 1462
University of Pennsylvania.  Fax: (215) 573 - 5325 or (215) 573 - 5315
**************************************************************************

 
 
 

SCSI command timeout

Post by Michael Haar » Fri, 03 Mar 1995 01:25:07


: I am also getting a similar problem. Our backups are run automatically
: overnight, using a combination of 'cpio' and 'dd'. At random, the backups
: would fail, leaving messages such as:


:    Disconnected command timeout for Target 4.0

:    SCSI transport failed: reason 'timeout'

: The tape drive (Sun 5Gb 4mm DAT), would then be left with one of the LEDs
: flashing.

I am having exactly the same problem.  It used to work fine in the past,
but now it fails _always_.  This means that we are currently unable to
make backups at all!!!


        Disconnected command timeout for Target 4.0
WARNING:

        SCSI transport failed: reason 'timeout': giving up

The machine in question is SPARC classic with a DAT drive.  The following
patches are applied:

 101316-01 101317-04 101317-09 101317-10 101318-36 101318-59 101327-02
 101327-08 101329-09 101331-03 101331-04 101331-05 101344-05 101344-11
 101347-01 101347-02 101362-09 101378-10 101415-02 101533-03 101533-05
 101615-01 101615-02 101739-04 101889-01 101889-03 102034-01 102167-01

Michael

 
 
 

SCSI command timeout

Post by James_Mathies » Wed, 08 Mar 1995 04:21:54




>I use ufsdump for backups, also automated to run at night. I never got the
>'timeout' warning at all, but ofthen I get the following message:

>WARNING:

>    Soft error rate (retries = 1) during writing was too high
>    Periodic head cleaning required and/or replace tape cartridge

I got this error too... on a brand new drive with a brand new tape.
We're using Sun's DAT 4mm 5gig component.  Anyone else seeing these
errors?

james

 
 
 

SCSI command timeout

Post by Andreas Kle » Thu, 16 Mar 1995 04:18:42


: I use ufsdump for backups, also automated to run at night. I never got the
: 'timeout' warning at all, but ofthen I get the following message:

: WARNING:

:       Soft error rate (retries = 1) during writing was too high
:       Periodic head cleaning required and/or replace tape cartridge

: This warning comes even after the head was cleaned with a cleaning cartridge the
: before inserting the backup tape!

Then you haven't installed your Solaris 2.4 Hardware 11/94 properly.
According to the Installation notes you shouldn't reboot after
installation. Instead of this

        cd /cdrom/Patches
        ./installpatch -u /a  (if I remember right)

This installs over 30 patches. Among them a patch that replaces
the tape driver. It is a bug in the tape driver that reports to
often, that the heads should be cleaned....

The Patch should also be available for 2.3 I think. Ask Sun for a
Patch or look onto your System CD or look at the Sunsolve Patch CD.
Or ask Sun if you still have warranty.

        Andreas ///

--


                  *** apsfilter - irgendwie clever ***
ftp.informatik.rwth-aachen.de:/pub/Linux/local/packs/APSfilter/aps-49...:-)

 
 
 

SCSI command timeout

Post by Dr. UN » Fri, 17 Mar 1995 00:11:15




: >
: >I use ufsdump for backups, also automated to run at night. I never got the
: >'timeout' warning at all, but ofthen I get the following message:
: >
: >WARNING:

: >  Soft error rate (retries = 1) during writing was too high
: >  Periodic head cleaning required and/or replace tape cartridge
: >

: I got this error too... on a brand new drive with a brand new tape.
: We're using Sun's DAT 4mm 5gig component.  Anyone else seeing these
: errors?

: james

 
 
 

SCSI command timeout

Post by Ross Mar » Sat, 18 Mar 1995 08:42:07




> Newsgroups: comp.unix.solaris,comp.sys.sun.hardware,comp.sys.sun.admin
> Followup-To: comp.unix.solaris,comp.sys.sun.hardware,comp.sys.sun.admin
> Date: 14 Mar 1995 19:18:42 GMT
> Organization: Solaris 2.4 research kitchen, D-41469 Neuss


> : I use ufsdump for backups, also automated to run at night. I never got the
> : 'timeout' warning at all, but ofthen I get the following message:

> : WARNING:

> :  Soft error rate (retries = 1) during writing was too high
> :  Periodic head cleaning required and/or replace tape cartridge

> : This warning comes even after the head was cleaned with a cleaning cartridge the
> : before inserting the backup tape!

> Then you haven't installed your Solaris 2.4 Hardware 11/94 properly.
> According to the Installation notes you shouldn't reboot after
> installation. Instead of this

>    cd /cdrom/Patches
>    ./installpatch -u /a  (if I remember right)

> This installs over 30 patches. Among them a patch that replaces
> the tape driver. It is a bug in the tape driver that reports to
> often, that the heads should be cleaned....

It was not a bug in the driver.  It is a bug in the tape drive firmware
where a busy scsi bus can incorrectly increase the error count.  The fix in
the driver was to help suppress this warning by increasing the error
allowance.

Quote:> The Patch should also be available for 2.3 I think. Ask Sun for a
> Patch or look onto your System CD or look at the Sunsolve Patch CD.
> Or ask Sun if you still have warranty.

There is no driver patch for 2.3.

------------------------------------------------------------------------

Sun Microsystems, Inc         |  sun!rambo
2550 Garcia Ave, UMPK-12-308  |  Voice: 415-786-6177
Mountain View, CA 94043       |  FAX  : 415-786-6437
------------------------------------------------------------------------
--
------------------------------------------------------------------------

Sun Microsystems, Inc         |  sun!rambo
2550 Garcia Ave, UMPK-12-308  |  
Mountain View, CA 94043       |  
------------------------------------------------------------------------

 
 
 

1. SCSI command timeout

I saw a similar problem on our HP server.  I tried using a Maxell 120s
tape, and got the error.  The system works fine with Maxell 90s tapes, but
we successfully use the other tapes in other DAT drives.

   ========================================================================


   |                                         |  I'll have wings."         |
   ========================================================================

2. SCO Open Server and Cyrix Processor?

3. SCSI command timeouts - Can't install

4. mod_ssl and apache client authentication

5. Default SCSI command Timeout calue??

6. Mail problem?

7. Help needed ! (scsi : aborting command due to timeout)

8. how do I determine which version of slackware I have install

9. scsi: aborting command due to timeout

10. SCSI timeouts during boot - command to ignore?

11. SCSI: aborting command due to timeout....?

12. SCSI: aborting command due to timeout

13. SCSI: Aborting command due to TIMEOUT