[stgt] VTL tape not working

[Albert Pauw]

After some time of absence I thought I gave the VTL stuff of tgt a try 
again.

Using FC9 and the latest (git) versions of open-iscsi and tgt I am able 
to see the scsi tape drive
with the following tgt settings:

<target iqn.2008-09.com.example:server.tape>
   backing-store /home/tapes/A3678299
#    incominguser someuser secretpass12
#    outgoinguser userA secretpassA
#    write-cache off
   device-type tape
   lun 1
   MaxRecvDataSegmentLength 8192
   MaxXmitDataSegmentLength 8192
   HeaderDigest CRC32C
   DataDigest None
   InitialR2T Yes
   MaxOutstandingR2T 1
   ImmediateData Yes
   FirstBurstLength 65536
   MaxBurstLength 262144
   DataPDUInOrder Yes
   DataSequenceInOrder Yes
   ErrorRecoveryLevel 0
   IFMarker No
   OFMarker No
   DefaultTime2Wait 2
   DefaultTime2Retain 20
   OFMarkInt Reject
   IFMarkInt Reject
   MaxConnections 1
</target>

I used the mktape command like this:

mktape -m A3678299 -s 10 -t data

Unfortunately, I only seem  to be able to store 10k of data in it:

-rw-r----- 1 root root 10240 2009-02-15 11:35 A3678299

After which I get scsi errors:

Feb 15 11:34:49 orange kernel: scsi3 : iSCSI Initiator over TCP/IP
Feb 15 11:34:49 orange kernel: scsi 3:0:0:0: RAID              IET      
Controller       0001 PQ: 0 ANSI: 5
Feb 15 11:34:49 orange kernel: scsi 3:0:0:0: Attached scsi generic sg2 
type 12
Feb 15 11:34:49 orange kernel: scsi 3:0:0:1: Sequential-Access IET      
VIRTUAL-TAPE     0001 PQ: 0 ANSI: 5
Feb 15 11:34:49 orange kernel: st 3:0:0:1: Attached scsi tape st0
Feb 15 11:34:49 orange kernel: st 3:0:0:1: st0: try direct i/o: yes 
(alignment 1 B)
Feb 15 11:34:49 orange kernel: st 3:0:0:1: Attached scsi generic sg3 type 1
Feb 15 11:34:49 orange iscsid: connection1:0 is operational now
Feb 15 11:35:25 orange kernel: st0: Block limits 4 - 1048576 bytes.
Feb 15 11:35:25 orange kernel: tgtd[15291]: segfault at 4 ip 08054468 sp 
ad60a320 error 4 in tgtd[8047000+25000]
Feb 15 11:35:25 orange tgtd: ssc_read_block_limit(108) In 
ssc_read_block_limit
Feb 15 11:35:30 orange kernel: connection1:0: detected conn error (1011)
Feb 15 11:35:30 orange kernel: st0: Error 20000 (sugg. bt 0x0, driver bt 
0x0, host bt 0x2).
Feb 15 11:35:30 orange iscsid: Kernel reported iSCSI connection 1:0 
error (1011) state (3)
Feb 15 11:36:39 orange iscsid: send fail Connection timed out
Feb 15 11:37:30 orange kernel: session1: session recovery timed out 
after 120 secs
Feb 15 11:37:30 orange kernel: st0: Error 10000 (sugg. bt 0x0, driver bt 
0x0, host bt 0x1).
Feb 15 11:37:30 orange kernel: st0: Error on write filemark.
Feb 15 11:37:30 orange kernel: st0: Error 10000 (sugg. bt 0x0, driver bt 
0x0, host bt 0x1).

Is this a know bug,  is the code not finished yet, or am I doing 
something wrong here?

Albert


-------------- next part --------------
An embedded message was scrubbed...
From: Albert Pauw <albert.pauw at gmail.com>
Subject: VTL tape not working
Date: Sun, 15 Feb 2009 11:44:03 +0100
Size: 5637
URL: <http://lists.wpkg.org/pipermail/stgt/attachments/20090215/71478b9d/attachment-0001.eml>