Virtual TapeServer 8.0 Release Notes

Troubleshooting | 57
 VTD with no connection to a logical tape is still presented on the bus. If a VTD is
created and connected to a logical tape then the VTD is disconnected, the VTD is still
shown to the client. An SG device is created and an Inquiry command succeeds. However,
a Status command does not and the VTD cannot actually be used. (TL-2324)
 I/O fails to resume after Fibre Channel cable is disconnected and HBA times out.
If you disconnect the Fibre Channel cable to the VTS for a time period that exceeds the
VTS FC HBA driver timeout value, I/O is interrupted and will not recover. (TL-1886)
WORKAROUND: Rediscover the Fibre Channel devices on the initiator or reboot the
initiator.
General operations
 Performance degrades when a disk fails because disk cache writes are disabled.
When a disk fails within a RAID set, the controller will automatically turn off disk cache
writes. This will decrease system performance because of this degraded condition. After
the failed disk is replaced and the RAID set is rebuilt, performance will return to normal
without need of intervention. Contact Support for field replacement instructions.
(TL-4513)
 Kernel call trace is displayed on console if GFS data is included in a
troubleshooting package. When generating a troubleshooting package, if the “Include
GFS data” checkbox is selected, a kernel call trace will be displayed on the console. This is
normal behavior and does not affect system operations. (TL-5111)
 Rerun job in process if power is lost to system. If power is lost while an export is in-
progress, re-run the job to ensure that data is written to storage as expected. (TL-4607)
 BRMS may not update correctly if disconnected libraries are present in BRMS.
BRMS is not updated correctly if disconnected libraries exist. (TL-3953)
WORKAROUND: Remove old non-responsive library devices and media from BRMS.
 GET FW STATE messages displayed on the console after restarting TapeServer.
Stopping and restarting TapeServer causes repeated messages similar to the following on
the console. (TL-3853)
Nov 15 13:46:52 vts15 kernel: isp0: Firmware State <Config Wait->Loss
Of Sync>
Nov 15 13:46:52 vts15 kernel: isp0: Interrupting Mailbox Command
(0x69) Timeout (elapsed time 100 jiffies)
Nov 15 13:46:52 vts15 kernel: isp0: Mailbox Command 'GET FW STATE'
failed (TIMEOUT)
Nov 15 13:46:52 vts15 kernel: isp0: Interrupting Mailbox Command
(0x69) Timeout (elapsed time 100 jiffies)
 When encrypting or decrypting a cartridge and TapeServer stops, temporary
directory and file are not deleted. Do not stop TapeServer or restart VTS unless the
system is quiesced. (TL-3071)
 SCSI EJECT command on an empty VTD hangs the calling client's mt command.
When sending a SCSI EJECT command (mt -f /dev/st? eject) to an empty VTD, the
command will take up to a minute to timeout prior to returning. In releases before 6.04,
this command would return immediately. (TL-2317)