🚀 go-pugleaf

Hello all,

I have a problem with a SCSI tape drive with which I'd appreciate some
help.  I can write to the drive but not read from it.  I've tried three
different tapes all with the same result. 

The drive works OK under DOS using Sytos Plus V1.21 backup software.  I
have successfully backed up and verified about 5MB of data, so at least I 
have some confidence that the hardware is OK. 

The hardware is an Archive Viper 2150S drive, Adaptec 1522 controller
(SoundBlaster 16 SCSI card).  Also on the SCSI bus is an NEC CD-ROM, it
works fine.  The bus is terminated correctly, terminators installed on the
controller and the CD, but no terminators on the tape drive.  The
controller and CD are at either end of the cable, the tape drive in
between.

I'm running Linux version 1.2.13 (RedHat 3.0.3) with the SCSI tape driver
compiled with debug on to try to get some useful information.

When I first started trying to write to the tape, both taper and tar 
would hang.  I later discovered that this was caused by the block size 
being incorrect, so now I'm using a block size of 512 bytes, this at 
least allow me to write to the tape (thanks to Andrew at the last SLUG 
meeting who suggested a smaller block size).

This is what mt tells me about the drive:

  $ mt -f /dev/st0 status
  SCSI 1 tape drive, residual=0
  File number=0, block number=0.
  Tape block size 512 bytes. Density code 0x10 (QIC-150/250 (GCR 10000 bpi)).
  Soft error count since last status=0
  General status bits on (41000000):
   BOT ONLINE


These are the commands I've been using to write to and read from the tape
and the results of each command:

  $ tar -cvf /dev/st0 -b 1 movemail-src/
  movemail-src/
  movemail-src/movemail.c
  movemail-src/COPYING
  movemail-src/README

This seems to work fine, the tape moves, but of course I can't verify that 
data is actually written to the tape.


  $ tar -b 1 -tvf /dev/st0

Then tar hangs, with the LED on the tape drive on. Examining
/var/log/messages shows that the SCSI command timed out. Eventually, the
tape drive is reset, the LED goes out, and tar gives the following error
message: 

  tar: Read error on /dev/st0: I/O error
  tar: At beginning of tape, quitting now


This is what's in /var/log/messages as a result of the second tar command
(lines wrapped for readability):

  Sep 10 22:21:31 localhost kernel: st0: Block limits 512 - 512 bytes.
  Sep 10 22:21:31 localhost kernel: st0: Mode sense. Length 12, medium 0, 
    WBS 10, BLL 8
  Sep 10 22:21:31 localhost kernel: st0: Density 10, tape length: 0, drv
    buffer: 1
  Sep 10 22:21:31 localhost kernel: st0: Block size: 512, buffer size: 32768 
    (64 blocks).
  Sep 10 22:36:31 localhost kernel: scsi : aborting command due to timeout : 
    pid 87, scsi0, id 4, lun 0 0x08 01 00 00 40 00 
  Sep 10 22:36:31 localhost kernel: aha152x: abort(), SCpnt=0x00232950,
    QUEUE STATUS:
  Sep 10 22:36:31 localhost kernel: issue_SC:
  Sep 10 22:36:31 localhost kernel: current_SC:
  Sep 10 22:36:31 localhost kernel: 0x00232950: target=4; lun=0; cmnd=(
    0x08 01 00 00 40 00 
  Sep 10 22:36:31 localhost kernel: ); residual=32768; buffers=0; phase 
    |send_ident|; in other(MESSAGE IN); next=0x00000000
  Sep 10 22:36:31 localhost kernel: disconnected_SC:
  Sep 10 22:36:31 localhost kernel: 
  Sep 10 22:36:31 localhost kernel: on bus: SCSISEQ ( ); SCSISIG ( MESSAGE
    IN ); INTSTAT ( lo ); SSTAT ( SELDI PHASECHG REQINIT ); SSTAT ( );
    SXFRCTL0 ( CH1 ); SIGNAL ( BSYI REQI ); SELID ( 90 ), SSTAT2 ( SEMPTY );
    SFCNT ( 0 ); FCNT ( 0 ); DMACNTRL0 ( 16BIT PIO WRITE INTEN ); 
  Sep 10 22:36:31 localhost kernel: enabled interrupts ( ENPHASEMIS
    ENBUSFREE )
  Sep 10 22:51:32 localhost kernel: scsi : aborting command due to timeout : 
    pid 87, scsi0, id 4, lun 0 0x08 01 00 00 40 00 
  Sep 10 22:51:32 localhost kernel: aha152x: abort(), SCpnt=0x00232950, 
    QUEUE STATUS:
  Sep 10 22:51:32 localhost kernel: issue_SC:
  Sep 10 22:51:32 localhost kernel: current_SC:
  Sep 10 22:51:32 localhost kernel: 0x00232950: target=4; lun=0; cmnd=(
    0x08 01 00 00 40 00 
  Sep 10 22:51:32 localhost kernel: ); residual=32768; buffers=0; phase 
    |send_ident|; in other(MESSAGE IN); next=0x00000000
  Sep 10 22:51:32 localhost kernel: disconnected_SC:
  Sep 10 22:51:32 localhost kernel: 
  Sep 10 22:51:32 localhost kernel: on bus: SCSISEQ ( ); SCSISIG ( MESSAGE
    IN ); INTSTAT ( lo ); SSTAT ( SELDI PHASECHG REQINIT ); SSTAT ( );
    SXFRCTL0 ( CH1 ); SIGNAL ( BSYI REQI ); SELID ( 90 ), SSTAT2 ( SEMPTY );
    SFCNT ( 0 ); FCNT ( 0 ); DMACNTRL0 ( 16BIT PIO WRITE INTEN ); 
  Sep 10 22:51:32 localhost kernel: enabled interrupts ( ENPHASEMIS
    ENBUSFREE )
  Sep 10 22:51:32 localhost kernel: SCSI host 0 abort() timed out - resetting
  Sep 10 22:51:32 localhost kernel: aha152x: reset(), bus not free: SCSI
    RESET OUT
  Sep 10 22:51:32 localhost kernel: QUEUE STATUS:
  Sep 10 22:51:32 localhost kernel: issue_SC:
  Sep 10 22:51:32 localhost kernel: current_SC:
  Sep 10 22:51:32 localhost kernel: 0x00232950: target=4; lun=0; cmnd=(
    0x08 01 00 00 40 00 
  Sep 10 22:51:32 localhost kernel: ); residual=32768; buffers=0; phase 
    |send_ident|; in other(MESSAGE IN); next=0x00000000
  Sep 10 22:51:32 localhost kernel: disconnected_SC:
  Sep 10 22:51:32 localhost kernel: 
  Sep 10 22:51:32 localhost kernel: on bus: SCSISEQ ( ); SCSISIG ( MESSAGE
    IN ); INTSTAT ( lo ); SSTAT ( SELDI PHASEMIS REQINIT ); SSTAT ( );
    SXFRCTL0 ( CH1 ); SIGNAL ( BSYI REQI ); SELID ( 90 ), SSTAT2 ( SEMPTY );
    SFCNT ( 0 ); FCNT ( 0 ); DMACNTRL0 ( 16BIT PIO READ ); 
  Sep 10 22:51:32 localhost kernel: enabled interrupts ( )
  Sep 10 22:51:32 localhost kernel: st0: Error: 80000, cmd: 8 1 0 0 40 0
    Len: 32768
  Sep 10 22:51:32 localhost kernel: st0: Error 80000. 
  Sep 10 22:51:32 localhost kernel: commands on targets w/ soft-resets:
  Sep 10 22:51:32 localhost kernel: QUEUE STATUS:
  Sep 10 22:51:32 localhost kernel: issue_SC:
  Sep 10 22:51:32 localhost kernel: current_SC:
  Sep 10 22:51:32 localhost kernel: none
  Sep 10 22:51:32 localhost kernel: disconnected_SC:
  Sep 10 22:51:32 localhost kernel: 
  Sep 10 22:51:32 localhost kernel: waiting: SCSISEQ ( ); SCSISIG ( MESSAGE 
    IN ); INTSTAT ( lo ); SSTAT ( SELDI PHASEMIS REQINIT ); SSTAT ( ); 
    SXFRCTL0 ( CH1 ); SIGNAL ( BSYI REQI ); SELID ( 90 ), SSTAT2 ( SEMPTY ); 
    SFCNT ( 0 ); FCNT ( 0 ); DMACNTRL0 ( 16BI T PIO READ ); 
  Sep 10 22:51:32 localhost kernel: enabled interrupts ( )
  Sep 10 22:51:32 localhost kernel: st0: Sense:  0  0  0  0  0  0  0  0
  Sep 10 22:51:33 localhost kernel: st0: Rewinding tape.
  Sep 10 22:51:33 localhost kernel: scsi : aborting command due to timeout : 
    pid 88, scsi0, id 4, lun 0 0x01 00 00 00 00 00 
  Sep 10 22:51:33 localhost kernel: aha152x: abort(), SCpnt=0x00232950,
    QUEUE STATUS:
  Sep 10 22:51:33 localhost kernel: issue_SC:
  Sep 10 22:51:33 localhost kernel: current_SC:
  Sep 10 22:51:33 localhost kernel: none
  Sep 10 22:51:33 localhost kernel: disconnected_SC:
  Sep 10 22:51:33 localhost kernel: 
  Sep 10 22:51:33 localhost kernel: waiting: SCSISEQ ( ); SCSISIG ( DATA
    OUT ); INTSTAT ( lo ); SSTAT ( SCSIRSTI BUSFREE ); SSTAT ( );
    SXFRCTL0 ( CH1 ); SIGNAL ( ); SELID ( 90 ), SSTAT2 ( SEMPTY );
    SFCNT ( 0 ); FCNT ( 0 ); DMACNTRL0 ( 16BIT PIO READ INTEN ) ; 
  Sep 10 22:51:33 localhost kernel: enabled interrupts ( )
  Sep 10 22:51:33 localhost kernel: command not found
  Sep 10 22:51:33 localhost kernel: aha152x: RESET IN
  Sep 10 22:51:33 localhost kernel: st0: Error: 28000000, cmd: 1 0 0 0 0 0
    Len: 16
  Sep 10 22:51:33 localhost kernel: extra data not valid Current error st0: 
    sns = 70  6
  Sep 10 22:51:33 localhost kernel: Raw sense data:0x70 0x00 0x06 0x00 0x00 
    0x00 0x00 0x00 


Anyone have any suggestions?

TIA,

John Clarke

-- 
jclarke@mpx.com.au                            jclarke@keycorp.com.au
              Windows 95: You make a grown man cry
"Master, does Emacs have the Buddha nature?" the novice asked.
"I don't see why not.  It's got bloody well everything else."