changes in tape-handling with 3.10.0-7?

udo

Distinguished Member
Apr 22, 2009
5,975
196
163
Ahrensburg; Germany
Hi,
since last week we changed our openvz-backup CT to kvm to be able to use the new kernel.
The new backup-VM works fine (bareos instead of bacula before) but the LTO-6 Jukebox must be connected to the host. Not an real problem.

But after changing the kernel on the host from 2.6.32-x to 3.10.0-7 I'm unable to mount tapes. Bareos see the tape but can't read the content.
Code:
3305 Autochanger "load slot 9, drive 1", status is OK.
block.c:1004 Read error on fd=4 at file:blk 0:0 on device "LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst). ERR=Das Gerät oder die Ressource ist belegt.
3902 Cannot mount Volume on Storage Device ""LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst)" because:
Requested Volume "" on "LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst) is not a Bareos labeled Volume, because: ERR=block.c:1004 Read error on fd=4 at file:blk 0:0 on device "LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst). ERR=Das Gerät oder die Ressource ist belegt.

Are there any known changes (e.g. block-size) which can produce such an issue? (we use 1MB-blocksize)

Udo
 
Hi,
since last week we changed our openvz-backup CT to kvm to be able to use the new kernel.
The new backup-VM works fine (bareos instead of bacula before) but the LTO-6 Jukebox must be connected to the host. Not an real problem.

But after changing the kernel on the host from 2.6.32-x to 3.10.0-7 I'm unable to mount tapes. Bareos see the tape but can't read the content.
Code:
3305 Autochanger "load slot 9, drive 1", status is OK.
block.c:1004 Read error on fd=4 at file:blk 0:0 on device "LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst). ERR=Das Gerät oder die Ressource ist belegt.
3902 Cannot mount Volume on Storage Device ""LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst)" because:
Requested Volume "" on "LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst) is not a Bareos labeled Volume, because: ERR=block.c:1004 Read error on fd=4 at file:blk 0:0 on device "LTO6-DRIVE2" (/dev/tape/by-path/pci-0000:0c:00.0-sas-0x5000e11155b61005-lun-0-nst). ERR=Das Gerät oder die Ressource ist belegt.

Are there any known changes (e.g. block-size) which can produce such an issue? (we use 1MB-blocksize)

Udo
Hi again,
the only different I have seen in the boot-log is one thing:

2.6.32:
Code:
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: Sequential-Access IBM      ULTRIUM-HH6      D8E5 PQ: 0 ANSI: 6
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: SSP: handle(0x0009), sas_addr(0x5000e11155b6100b), phy(4), device_name(0x5000e11155b6100a)
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: SSP: enclosure_logical_id(0x500605b008d9e610), slot(4)
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: qdepth(254), tagged(1), [B]simple(1)[/B], ordered(0), scsi_level(7), cmd_que(1)
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: TLR Enabled
3.10:
Code:
Mar 12 16:19:12 proxmox3 kernel: [    1.178296] scsi 2:0:0:0: Sequential-Access IBM      ULTRIUM-HH6      D8E5 PQ: 0 ANSI: 6
Mar 12 16:19:12 proxmox3 kernel: [    1.178307] scsi 2:0:0:0: SSP: handle(0x0009), sas_addr(0x5000e11155b6100b), phy(4), device_name(0x5000e11155b6100a)
Mar 12 16:19:12 proxmox3 kernel: [    1.178309] scsi 2:0:0:0: SSP: enclosure_logical_id(0x500605b008d9e610), slot(4)
Mar 12 16:19:12 proxmox3 kernel: [    1.178312] scsi 2:0:0:0: qdepth(254), tagged(1), [B]simple(0)[/B], ordered(0), scsi_level(7), cmd_que(1)
Mar 12 16:19:12 proxmox3 kernel: [    1.180553] scsi 2:0:0:0: TLR Enabled
Don't know what simple changed...

Udo
 
Hi,
Did you find out after that the root and the solution of the problem ? It seems we have a similar one.
Thanks,

Hi again,
the only different I have seen in the boot-log is one thing:

2.6.32:
Code:
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: Sequential-Access IBM      ULTRIUM-HH6      D8E5 PQ: 0 ANSI: 6
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: SSP: handle(0x0009), sas_addr(0x5000e11155b6100b), phy(4), device_name(0x5000e11155b6100a)
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: SSP: enclosure_logical_id(0x500605b008d9e610), slot(4)
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: qdepth(254), tagged(1), [B]simple(1)[/B], ordered(0), scsi_level(7), cmd_que(1)
Mar 13 11:48:56 proxmox3 kernel: scsi 2:0:0:0: TLR Enabled
3.10:
Code:
Mar 12 16:19:12 proxmox3 kernel: [    1.178296] scsi 2:0:0:0: Sequential-Access IBM      ULTRIUM-HH6      D8E5 PQ: 0 ANSI: 6
Mar 12 16:19:12 proxmox3 kernel: [    1.178307] scsi 2:0:0:0: SSP: handle(0x0009), sas_addr(0x5000e11155b6100b), phy(4), device_name(0x5000e11155b6100a)
Mar 12 16:19:12 proxmox3 kernel: [    1.178309] scsi 2:0:0:0: SSP: enclosure_logical_id(0x500605b008d9e610), slot(4)
Mar 12 16:19:12 proxmox3 kernel: [    1.178312] scsi 2:0:0:0: qdepth(254), tagged(1), [B]simple(0)[/B], ordered(0), scsi_level(7), cmd_que(1)
Mar 12 16:19:12 proxmox3 kernel: [    1.180553] scsi 2:0:0:0: TLR Enabled
Don't know what simple changed...

Udo
 
Hi,
Did you find out after that the root and the solution of the problem ? It seems we have a similar one.
Thanks,
Hi,
no - the "old" proxmox-server is still on 2.6.32...

But I have installed a new one for an dependance with pve4 (4.2.3-2-pve) and the jukebox access work well.
One remark: the backup software is in this case bareos 15.2 instead of 14.2 on the "old" server.

The message output is slightly different:
Code:
[    3.072811] scsi 1:0:0:0: Sequential-Access IBM      ULT3580-HH6      E6R3 PQ: 0 ANSI: 6
[    3.072816] scsi 1:0:0:0: SSP: handle(0x0009), sas_addr(0x5000e1115d398002), phy(3), device_name(0x5000e1115d398001)
[    3.072818] scsi 1:0:0:0: SSP: enclosure_logical_id(0x500605b00a4a7c20), slot(3)
[    3.075919] scsi 1:0:0:0: TLR Enabled
[    3.083605] scsi 1:0:0:0: Attached scsi generic sg2 type 1
[    3.085452] st 1:0:0:0: Attached scsi tape st0
Udo
 
OK ; but what is the version of the kernel on your "new server" ? You explained that what made the things go wrong was upgrading the kernel, isn't it ? So we could expect that is the version of kernel that has the most impacts for this problem...

Hi,
no - the "old" proxmox-server is still on 2.6.32...

But I have installed a new one for an dependance with pve4 (4.2.3-2-pve) and the jukebox access work well.
One remark: the backup software is in this case bareos 15.2 instead of 14.2 on the "old" server.

The message output is slightly different:
Code:
[    3.072811] scsi 1:0:0:0: Sequential-Access IBM      ULT3580-HH6      E6R3 PQ: 0 ANSI: 6
[    3.072816] scsi 1:0:0:0: SSP: handle(0x0009), sas_addr(0x5000e1115d398002), phy(3), device_name(0x5000e1115d398001)
[    3.072818] scsi 1:0:0:0: SSP: enclosure_logical_id(0x500605b00a4a7c20), slot(3)
[    3.075919] scsi 1:0:0:0: TLR Enabled
[    3.083605] scsi 1:0:0:0: Attached scsi generic sg2 type 1
[    3.085452] st 1:0:0:0: Attached scsi tape st0
Udo
 

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!