I'm trying to use an old Seagate Travan SCSI tape drive in my Fedora Core 2 system to restore some data off old backup tapes and I'm not having any luck.
Kudzu found the SCSI card (Adaptec 2940) and it shows up in lscpi: 00:0a.0 SCSI storage controller: Adaptec AHA-2940/2940W / AIC-7871
The modprobe.conf now contains the following line: alias scsi_hostadapter aic7xxx
I did a Google search and someone suggested trying to manually load aic7xxx_old instead. That seemed to get the tape drive to get recognized but then something weird happens and I still can't access the tape drive.
This is what was found in /var/log/messages when using aic7xxxx:
Aug 27 10:46:03 penguin kernel: (scsi4:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Aug 27 10:46:03 penguin kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.17 Aug 27 10:46:03 penguin kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 27 10:46:04 penguin kernel: Attached scsi tape st0 at scsi4, channel 0, id 4, lun 0 Aug 27 10:46:04 penguin kernel: st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Aug 27 10:46:04 penguin kernel: Attached scsi generic sg5 at scsi4, channel 0, id 4, lun 0, type 1 Aug 27 10:46:04 penguin scsi.agent[6435]: tape at /devices/pci0000:00/0000:00:0a.0/host4/4:0:4:0 Aug 27 10:46:13 penguin kernel: (scsi4:A:4:0): parity error detected in Data-in phase. SEQADDR(0x1b5) SCSIRATE(0x4f) Aug 27 10:46:13 penguin kernel: st0: Error 70002 (sugg. bt 0x0, driver bt 0x0, host bt 0x7).
Here is what's in dmesg: Adaptec AHA-294X SCSI host adapter did not call scsi_unregister [<52baaade>] exit_this_scsi_driver+0xa7/0xec [aic7xxx_old] [<02136845>] sys_delete_module+0x129/0x170 [<02151db2>] unmap_vma_list+0xe/0x17 [<0215218a>] do_munmap+0x1d8/0x1e2 [<021181a7>] do_page_fault+0x0/0x489 ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi4 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi4:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi4, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi4, channel 0, id 4, lun 0, type 1 (scsi4:A:4:0): parity error detected in Data-in phase. SEQADDR(0x1b5) SCSIRATE(0x4f) st0: Error 70002 (sugg. bt 0x0, driver bt 0x0, host bt 0x7).
What does that "parity error" mean???
Gregory Gulik wrote:
I'm trying to use an old Seagate Travan SCSI tape drive in my Fedora Core 2 system to restore some data off old backup tapes and I'm not having any luck.
Kudzu found the SCSI card (Adaptec 2940) and it shows up in lscpi: 00:0a.0 SCSI storage controller: Adaptec AHA-2940/2940W / AIC-7871
The modprobe.conf now contains the following line: alias scsi_hostadapter aic7xxx
I did a Google search and someone suggested trying to manually load aic7xxx_old instead. That seemed to get the tape drive to get recognized but then something weird happens and I still can't access the tape drive.
This is what was found in /var/log/messages when using aic7xxxx:
Aug 27 10:46:03 penguin kernel: (scsi4:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Aug 27 10:46:03 penguin kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.17 Aug 27 10:46:03 penguin kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 27 10:46:04 penguin kernel: Attached scsi tape st0 at scsi4, channel 0, id 4, lun 0 Aug 27 10:46:04 penguin kernel: st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Aug 27 10:46:04 penguin kernel: Attached scsi generic sg5 at scsi4, channel 0, id 4, lun 0, type 1 Aug 27 10:46:04 penguin scsi.agent[6435]: tape at /devices/pci0000:00/0000:00:0a.0/host4/4:0:4:0 Aug 27 10:46:13 penguin kernel: (scsi4:A:4:0): parity error detected in Data-in phase. SEQADDR(0x1b5) SCSIRATE(0x4f) Aug 27 10:46:13 penguin kernel: st0: Error 70002 (sugg. bt 0x0, driver bt 0x0, host bt 0x7).
Here is what's in dmesg: Adaptec AHA-294X SCSI host adapter did not call scsi_unregister [<52baaade>] exit_this_scsi_driver+0xa7/0xec [aic7xxx_old] [<02136845>] sys_delete_module+0x129/0x170 [<02151db2>] unmap_vma_list+0xe/0x17 [<0215218a>] do_munmap+0x1d8/0x1e2 [<021181a7>] do_page_fault+0x0/0x489 ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi4 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi4:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi4, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi4, channel 0, id 4, lun 0, type 1 (scsi4:A:4:0): parity error detected in Data-in phase. SEQADDR(0x1b5) SCSIRATE(0x4f) st0: Error 70002 (sugg. bt 0x0, driver bt 0x0, host bt 0x7).
What does that "parity error" mean???
I'm getting something similar with a new FC2 install on a Dell Power Edge 750 server.
I have this working properly on an old clone running FC1, but it seems FC1 doesn't like the NICs that Dell put into the new server.
Tape drive is a Sony SDX-700C AIT3, and it's in a Breece-Hill 8 pack library, SCSI card is an Adaptec 29160. in FC2 I can do a cat /proc/scsi/scsi and it sees the SATA drive, the library and the AIT3.
Sony has a utility called "sonytape" however it DOES NOT recognize the drive. Neither does NetVault 7.1.
Ideas anyone?
On Fri, 2004-08-27 at 10:47, Gregory Gulik wrote:
Attached scsi generic sg5 at scsi4, channel 0, id 4, lun 0, type 1 (scsi4:A:4:0): parity error detected in Data-in phase. SEQADDR(0x1b5) SCSIRATE(0x4f) st0: Error 70002 (sugg. bt 0x0, driver bt 0x0, host bt 0x7).
What does that "parity error" mean???
It could really be an error. Is the scsi cable terminated correctly either on the cable or (only) the last device and if one or the other of the internal or external cables is unused the card should have it's terminator activated or on 'auto'. Or, this might be one of those devices where you have to disable parity in the card bios.
--- Les Mikesell les@futuresource.com
I first suspected the cable so I tried a couple SCSI cables. I didn't think of trying to turn off parity on the drive. I'll try that if I can figure out how to disable it in the BIOS.
Les Mikesell wrote:
It could really be an error. Is the scsi cable terminated correctly either on the cable or (only) the last device and if one or the other of the internal or external cables is unused the card should have it's terminator activated or on 'auto'. Or, this might be one of those devices where you have to disable parity in the card bios.
On Fri, Aug 27, 2004 at 10:47:20AM -0500, Gregory Gulik wrote:
I'm trying to use an old Seagate Travan SCSI tape drive in my Fedora Core 2 system to restore some data off old backup tapes and I'm not having any luck.
Kudzu found the SCSI card (Adaptec 2940) and it shows up in lscpi: 00:0a.0 SCSI storage controller: Adaptec AHA-2940/2940W / AIC-7871
The modprobe.conf now contains the following line: alias scsi_hostadapter aic7xxx
I did a Google search and someone suggested trying to manually load aic7xxx_old instead. That seemed to get the tape drive to get recognized but then something weird happens and I still can't access the tape drive.
This is what was found in /var/log/messages when using aic7xxxx:
Aug 27 10:46:03 penguin kernel: (scsi4:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Aug 27 10:46:03 penguin kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.17 Aug 27 10:46:03 penguin kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 27 10:46:04 penguin kernel: Attached scsi tape st0 at scsi4, channel 0, id 4, lun 0 Aug 27 10:46:04 penguin kernel: st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Aug 27 10:46:04 penguin kernel: Attached scsi generic sg5 at scsi4, channel 0, id 4, lun 0, type 1 Aug 27 10:46:04 penguin scsi.agent[6435]: tape at /devices/pci0000:00/0000:00:0a.0/host4/4:0:4:0 Aug 27 10:46:13 penguin kernel: (scsi4:A:4:0): parity error detected in Data-in phase. SEQADDR(0x1b5) SCSIRATE(0x4f) Aug 27 10:46:13 penguin kernel: st0: Error 70002 (sugg. bt 0x0, driver bt 0x0, host bt 0x7).
Do you have parity enabled or disabled consistently on that SCSI bus? I.e. all devices have parity enabled or all have it disabled, including the HA. Parity enable on this drive is jumpers 7 and 8, next to the three address bit jumpers. Parity enable in the 2940 is via the BIOS firmware.
Also, you have firmware rev 1.17. You might want to upgrade it. I seem to recall problems with 1.17. I have 1.22, and there may be even more recent firmware.
What does that "parity error" mean???
It's an optional data integrity check.
On Fri, Aug 27, 2004 at 12:34:33PM -0500, Gregory Gulik wrote:
I first suspected the cable so I tried a couple SCSI cables. I didn't think of trying to turn off parity on the drive. I'll try that if I can figure out how to disable it in the BIOS.
CTL-A as soon as you see the prompt during boot.
Some people turn off the prompt. If you don't see it, just hit CTL-A from time to time after the memory test.
Charles Curley wrote:
Do you have parity enabled or disabled consistently on that SCSI bus? I.e. all devices have parity enabled or all have it disabled, including the HA. Parity enable on this drive is jumpers 7 and 8, next to the three address bit jumpers. Parity enable in the 2940 is via the BIOS firmware.
Ok, the parity error is now fixed. I had parity enabled in the BIOS and disabled on the drive. The parity error is now gone.
However I still can't get to the tape drive. I tried both the new and old drivers.
Aug 27 12:57:31 penguin kernel: ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 Aug 27 12:57:31 penguin kernel: scsi5 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 Aug 27 12:57:31 penguin kernel: <Adaptec 2940 SCSI adapter> Aug 27 12:57:31 penguin kernel: aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs Aug 27 12:57:31 penguin kernel: Aug 27 12:57:49 penguin kernel: (scsi5:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Aug 27 12:57:49 penguin kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.17 Aug 27 12:57:49 penguin kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 27 12:57:49 penguin kernel: Attached scsi tape st0 at scsi5, channel 0, id 4, lun 0 Aug 27 12:57:49 penguin kernel: st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Aug 27 12:57:49 penguin kernel: Attached scsi generic sg5 at scsi5, channel 0, id 4, lun 0, type 1
ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi5 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi5:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi5, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi5, channel 0, id 4, lun 0, type 1
Also, you have firmware rev 1.17. You might want to upgrade it. I seem to recall problems with 1.17. I have 1.22, and there may be even more recent firmware.
I'll try that next.
Now I remember why I avoid SCSI...
On Fri, 2004-08-27 at 12:34, Gregory Gulik wrote:
I first suspected the cable so I tried a couple SCSI cables. I didn't think of trying to turn off parity on the drive. I'll try that if I can figure out how to disable it in the BIOS.
Termination could be the problem too if you haven't checked it carefully. Newer systems put terminators on the ends of the cables and the corresponding devices don't even have the option so people tend to forget about the issue. When you dig out an old device you need to make sure it doesn't have the terminator jumper set if you use a new terminated cable or if you use an old cable you need to do something to terminate the end or the last device. (Really old devices had resistor packs instead of jumpers...).
--- Les Mikesell les@futuresource.com
On Fri, Aug 27, 2004 at 12:58:47PM -0500, Gregory Gulik wrote:
Charles Curley wrote:
Do you have parity enabled or disabled consistently on that SCSI bus? I.e. all devices have parity enabled or all have it disabled, including the HA. Parity enable on this drive is jumpers 7 and 8, next to the three address bit jumpers. Parity enable in the 2940 is via the BIOS firmware.
Ok, the parity error is now fixed. I had parity enabled in the BIOS and disabled on the drive. The parity error is now gone.
Good...
However I still can't get to the tape drive. I tried both the new and old drivers.
What do you mean, you can't get to the tape drive? What have you tried that lead you to this conclusion? The log excerpts below indicate the driver is finding it and attaching it to /dev/st0.
You should see it with something like this:
[root@charlesc root]# mt -f /dev/st0 status SCSI 2 tape drive: File number=-1, block number=-1, partition=0. Tape block size 0 bytes. Density code 0x0 (default). Soft error count since last status=0 General status bits on (50000): DR_OPEN IM_REP_EN
Rather than something like this:
[root@charlesc root]# mt -f /dev/st3 status /dev/st3: No such device or address
Your drive should be at st0 if it's the first or only SCSI tape drive.
Aug 27 12:57:31 penguin kernel: ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 Aug 27 12:57:31 penguin kernel: scsi5 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 Aug 27 12:57:31 penguin kernel: <Adaptec 2940 SCSI adapter> Aug 27 12:57:31 penguin kernel: aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs Aug 27 12:57:31 penguin kernel: Aug 27 12:57:49 penguin kernel: (scsi5:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Aug 27 12:57:49 penguin kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.17 Aug 27 12:57:49 penguin kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 27 12:57:49 penguin kernel: Attached scsi tape st0 at scsi5, channel 0, id 4, lun 0 Aug 27 12:57:49 penguin kernel: st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Aug 27 12:57:49 penguin kernel: Attached scsi generic sg5 at scsi5, channel 0, id 4, lun 0, type 1
This looks like FC2 (2.6 kernel). Here's what my log looks like on FC1:
Aug 22 09:49:19 charlesc kernel: blk: queue dfeb5814, I/O limit 4095Mb (mask 0xffffffff) Aug 22 09:49:19 charlesc kernel: (scsi0:A:6): 6.944MB/s transfers (6.944MHz, offset 15) Aug 22 09:49:19 charlesc kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.22 Aug 22 09:49:19 charlesc kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 22 09:49:19 charlesc kernel: blk: queue df5f0614, I/O limit 4095Mb (mask 0xffffffff)
ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi5 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi5:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi5, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi5, channel 0, id 4, lun 0, type 1
Aug 22 09:49:19 charlesc kernel: scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 Aug 22 09:49:19 charlesc kernel: <Adaptec 2940A Ultra SCSI adapter> Aug 22 09:49:19 charlesc kernel: aic7860: Ultra Single Channel A, SCSI Id=7, 3/253 SCBs Aug 22 09:49:19 charlesc kernel: Aug 22 09:49:19 charlesc kernel: blk: queue df5f0414, I/O limit 4095Mb (mask 0xffffffff)
Also, you have firmware rev 1.17. You might want to upgrade it. I seem to recall problems with 1.17. I have 1.22, and there may be even more recent firmware.
I'll try that next.
Now I remember why I avoid SCSI...
-- Greg Gulik http://www.gulik.org/greg/ greg @ gulik.org
-- fedora-list mailing list fedora-list@redhat.com To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list
That's exactly it. The device is appearing to dmesg upon boot or modprobe for aic7xxx yet I can't access the device:
PCI: Enabling device 0000:00:0a.0 (0006 -> 0007) ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi6 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi6:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi6, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi6, channel 0, id 4, lun 0, type 1 [root@penguin root]# mt -f /dev/st0 status /dev/st0: Input/output error [root@penguin root]#
Charles Curley wrote:
What do you mean, you can't get to the tape drive? What have you tried that lead you to this conclusion? The log excerpts below indicate the driver is finding it and attaching it to /dev/st0.
You should see it with something like this:
[root@charlesc root]# mt -f /dev/st0 status SCSI 2 tape drive: File number=-1, block number=-1, partition=0. Tape block size 0 bytes. Density code 0x0 (default). Soft error count since last status=0 General status bits on (50000): DR_OPEN IM_REP_EN
Rather than something like this:
[root@charlesc root]# mt -f /dev/st3 status /dev/st3: No such device or address
Your drive should be at st0 if it's the first or only SCSI tape drive.
Aug 27 12:57:31 penguin kernel: ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 Aug 27 12:57:31 penguin kernel: scsi5 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 Aug 27 12:57:31 penguin kernel: <Adaptec 2940 SCSI adapter> Aug 27 12:57:31 penguin kernel: aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs Aug 27 12:57:31 penguin kernel: Aug 27 12:57:49 penguin kernel: (scsi5:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Aug 27 12:57:49 penguin kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.17 Aug 27 12:57:49 penguin kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 27 12:57:49 penguin kernel: Attached scsi tape st0 at scsi5, channel 0, id 4, lun 0 Aug 27 12:57:49 penguin kernel: st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Aug 27 12:57:49 penguin kernel: Attached scsi generic sg5 at scsi5, channel 0, id 4, lun 0, type 1
This looks like FC2 (2.6 kernel). Here's what my log looks like on FC1:
Aug 22 09:49:19 charlesc kernel: blk: queue dfeb5814, I/O limit 4095Mb (mask 0xffffffff) Aug 22 09:49:19 charlesc kernel: (scsi0:A:6): 6.944MB/s transfers (6.944MHz, offset 15) Aug 22 09:49:19 charlesc kernel: Vendor: CONNER Model: CTT8000-S Rev: 1.22 Aug 22 09:49:19 charlesc kernel: Type: Sequential-Access ANSI SCSI revision: 02 Aug 22 09:49:19 charlesc kernel: blk: queue df5f0614, I/O limit 4095Mb (mask 0xffffffff)
ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi5 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi5:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi5, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi5, channel 0, id 4, lun 0, type 1
Aug 22 09:49:19 charlesc kernel: scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 Aug 22 09:49:19 charlesc kernel: <Adaptec 2940A Ultra SCSI adapter> Aug 22 09:49:19 charlesc kernel: aic7860: Ultra Single Channel A, SCSI Id=7, 3/253 SCBs Aug 22 09:49:19 charlesc kernel: Aug 22 09:49:19 charlesc kernel: blk: queue df5f0414, I/O limit 4095Mb (mask 0xffffffff)
Also, you have firmware rev 1.17. You might want to upgrade it. I seem to recall problems with 1.17. I have 1.22, and there may be even more recent firmware.
I'll try that next.
Now I remember why I avoid SCSI...
-- Greg Gulik http://www.gulik.org/greg/ greg @ gulik.org
-- fedora-list mailing list fedora-list@redhat.com To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list
On Mon, 2004-08-30 at 14:37, Gregory Gulik wrote:
That's exactly it. The device is appearing to dmesg upon boot or modprobe for aic7xxx yet I can't access the device:
PCI: Enabling device 0000:00:0a.0 (0006 -> 0007) ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 scsi6 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940 SCSI adapter> aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs
(scsi6:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi tape st0 at scsi6, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 Attached scsi generic sg5 at scsi6, channel 0, id 4, lun 0, type 1 [root@penguin root]# mt -f /dev/st0 status /dev/st0: Input/output error [root@penguin root]#
It doesn't look like an issue since the driver is detecting the transfer rate of 5mb/s, but you could try setting that rate in the SCSI adapter for that SCSI ID. Also, you might want to see if you can get this tape drive working on another system.
For someone who avoids SCSI, you sure do have a lot of SCSI controllers in this machine.
C. Linus Hicks wrote:
It doesn't look like an issue since the driver is detecting the transfer rate of 5mb/s, but you could try setting that rate in the SCSI adapter for that SCSI ID. Also, you might want to see if you can get this tape drive working on another system.
I tried setting the transfer rate to 5mb/s in the BIOS. No dice. However I did notice the following error in dmesg after I tried to access the device with mt:
st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
For someone who avoids SCSI, you sure do have a lot of SCSI controllers in this machine.
The only real device is this Adaptec card. The rest are a 6-in-1 USB card reader and a large IDE drive in a USB enclosure as a backup device.
On Mon, 2004-08-30 at 22:37, Gregory Gulik wrote:
C. Linus Hicks wrote:
It doesn't look like an issue since the driver is detecting the transfer rate of 5mb/s, but you could try setting that rate in the SCSI adapter for that SCSI ID. Also, you might want to see if you can get this tape drive working on another system.
I tried setting the transfer rate to 5mb/s in the BIOS. No dice. However I did notice the following error in dmesg after I tried to access the device with mt:
st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
I don't think that is an error. I have seen it in every log listing you have posted.
For someone who avoids SCSI, you sure do have a lot of SCSI controllers in this machine.
The only real device is this Adaptec card. The rest are a 6-in-1 USB card reader and a large IDE drive in a USB enclosure as a backup device.
Ah! So do you have any other SCSI devices you can test on this card to verify that it works?
On Mon, 2004-08-30 at 22:21, C. Linus Hicks wrote:
On Mon, 2004-08-30 at 22:37, Gregory Gulik wrote:
C. Linus Hicks wrote:
It doesn't look like an issue since the driver is detecting the transfer rate of 5mb/s, but you could try setting that rate in the SCSI adapter for that SCSI ID. Also, you might want to see if you can get this tape drive working on another system.
I tried setting the transfer rate to 5mb/s in the BIOS. No dice. However I did notice the following error in dmesg after I tried to access the device with mt:
st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
I don't think that is an error. I have seen it in every log listing you have posted.
For someone who avoids SCSI, you sure do have a lot of SCSI controllers in this machine.
The only real device is this Adaptec card. The rest are a 6-in-1 USB card reader and a large IDE drive in a USB enclosure as a backup device.
Ah! So do you have any other SCSI devices you can test on this card to verify that it works? -- C. Linus Hicks lhicks@nc.rr.com
duh Oh!
He said he only has one scsi device. All the others he mentioned are USB (scsi emulation) and do not use the scsi adapter.
Dear Mates: I would like your advice on this matter: using Fedora core 2 and having difficulties upon automounting cdrom and cdrw drives. FSTAB file is listed below...
LABEL=/ / ext3 defaults 1 1 none /dev/pts devpts gid=5,mode=620 0 0 none /dev/shm tmpfs defaults 0 0 none /proc proc defaults 0 0 none /sys sysfs defaults 0 0 /dev/hda1 /win_c vfat defaults 0 0 /dev/hda5 /win_d vfat defaults 0 0 /dev/hdb2 swap swap defaults 0 0 /dev/cdrom /mnt/cdrom auto auto,user,exec,ro 0 0 /dev/cdrom1 /mnt/cdrom1 auto auto,user,exec,rw 0 0 /dev/fd0 /mnt/floppy auto auto,user,sync,rw 0 0
I have already edited it, so it is different from original one. It seems there is no problem to mount Linux-made cd's. Kudzu parameter and filesystem "udf,iso 9660" did not work either. Floppy disk and hda are mounted with no glitches. Thanks in advance,
I did locate an old SCSI CD-ROM drive and put it in. It's not being recognized at all. So either the problem is the card or the driver. Now I need to try to figure out which it is.
*SIGH*
Jeff Vian wrote:
duh Oh!
He said he only has one scsi device. All the others he mentioned are USB (scsi emulation) and do not use the scsi adapter.
On Thursday 02 September 2004 12:07, Gregory Gulik wrote:
I did locate an old SCSI CD-ROM drive and put it in. It's not being recognized at all. So either the problem is the card or the driver. Now I need to try to figure out which it is.
*SIGH*
Jeff Vian wrote:
duh Oh!
He said he only has one scsi device. All the others he mentioned are USB (scsi emulation) and do not use the scsi adapter.
-- Greg Gulik http://www.gulik.org/greg/ greg @ gulik.org
Coming into this thread late. Which Adaptec controller are you using? Also, what SCSI tape unit?
Tom
Gregory Gulik wrote:
Jeff Vian wrote:
duh Oh!
He said he only has one scsi device. All the others he mentioned are USB (scsi emulation) and do not use the scsi adapter.
I did locate an old SCSI CD-ROM drive and put it in. It's not being recognized at all. So either the problem is the card or the driver. Now I need to try to figure out which it is.
*SIGH*
Sounds like you're grasping for straws, so here's a straw:
I recently had a case of an Adaptec 2940 with tape and acanner attached doing some weird crap. The fix was popping the motherboard battery out of its socket and temporarily jumpering the pins that cause CMOS to be cleared. This was an ASUS A7N8X - deluxe Ver 2.0 m/b. YMMV
Thanks for the tip. I'm doing this on an ASUS A7V8X-X motherboard also!
Robert wrote:
Sounds like you're grasping for straws, so here's a straw:
I recently had a case of an Adaptec 2940 with tape and acanner attached doing some weird crap. The fix was popping the motherboard battery out of its socket and temporarily jumpering the pins that cause CMOS to be cleared. This was an ASUS A7N8X - deluxe Ver 2.0 m/b. YMMV
It's the Adaptec 2940 SCSI card with a Seagate Travan 10G/20G tape drive that identifies itself as a Conner CTT8000-S.
You can catch up on the thread in the archives:
http://www.redhat.com/archives/fedora-list/2004-August/msg05266.html
I tried resetting the CMOS to clear my BIOS settings and except for losing my mouse for a while it didn't do any good. As a matter of fact, it's worse than before:
Sep 3 11:10:26 penguin kernel: ACPI: PCI interrupt 0000:00:0a.0[A] -> GSI 11 (level, low) -> IRQ 11 Sep 3 11:10:26 penguin kernel: scsi3 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 Sep 3 11:10:26 penguin kernel: <Adaptec 2940 SCSI adapter> Sep 3 11:10:26 penguin kernel: aic7870: Single Channel A, SCSI Id=7, 16/253 SCBs Sep 3 11:10:26 penguin kernel: Sep 3 11:10:42 penguin kernel: (scsi3:A:4:0): Unexpected busfree in Message-out phase Sep 3 11:10:42 penguin kernel: SEQADDR == 0x165 Sep 3 11:10:42 penguin kernel: (scsi3:A:4:0): Unexpected busfree in Message-out phase Sep 3 11:10:42 penguin kernel: SEQADDR == 0x165 Sep 3 11:10:42 penguin kernel: (scsi3:A:4:0): Unexpected busfree in Message-out phase Sep 3 11:10:42 penguin kernel: SEQADDR == 0x165 Sep 3 11:10:42 penguin kernel: (scsi3:A:4:0): Unexpected busfree in Message-out phase Sep 3 11:10:42 penguin kernel: SEQADDR == 0x165 . . .
Eeek... Anyway, I don't have another SCSI card to try and the only other SCSI device I could find, a SCSI CD-ROM drive didn't work either. I'm making arrangements to borrow a friends SCSI card to try it out in case this card is just plain bad.
Tom Taylor wrote:
Coming into this thread late. Which Adaptec controller are you using? Also, what SCSI tape unit?
When the SCSI BIOS shows up on your boot screen, does it see the CD-ROM drive? If not, the most likely cause for it not beeing seen is that it is not working. Usually, you do not need a driver for SCSI devices - I have never needed one in my 11 years of Linux. Things to check: Does the CD-ROM drive have a unique SCSI-ID? This is settable with a few (3) jumpers and is binary encoded. No jumper -> SCSI-ID 0, 3 jumpers -> SCSI ID 7 - which is usually used by the Adaptor, so chose 0 - 6 only.
Do you have the SCSI termination right? The last device on the kable needs to be terminated or you have to add a terminator (not Arnold) to the last connector of the SCSI cable. You should also set the adaptor to be self terminating and make it the first/last device on the cable. Never have the adaptor sitting somewhere in the middle.
Best regards, Chris
On Thu, 2004-09-02 at 15:07, Gregory Gulik wrote:
I did locate an old SCSI CD-ROM drive and put it in. It's not being recognized at all. So either the problem is the card or the driver. Now I need to try to figure out which it is.
*SIGH*
Jeff Vian wrote:
duh Oh!
He said he only has one scsi device. All the others he mentioned are USB (scsi emulation) and do not use the scsi adapter.
-- Greg Gulik http://www.gulik.org/greg/ greg @ gulik.org
Sorry for the delay in getting back to everyone on this.
Anyway, quick update, my SCSI tape drive doesn't work. The original post was here:
http://www.redhat.com/archives/fedora-list/2004-August/msg05266.html
First, to answer the question below, yes, the CD-ROM drive and the tape drive are found by the SCSI card BIOS screen just fine. If I use the built-in utilities the device comes up just fine as well.
Anyway, I tried several things and something isn't right. My brother-in-law gave me his known working Adaptec 2940U card, same model as mine, just appears to be a newer revision. I also have a different internal SCSI cable from him, this one with a terminator built-in to one end.
After rebooting Kudzu found something changed and had me unconfigure the old card and configure the new card. Fine with me.
However when it finished booting I found the card wasn't really recognized. The aic7xxxx driver had not been loaded. WTF!
If I load it manually with "modprobe aic7xxx" it shows up in dmesg:
scsi3 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940A Ultra SCSI adapter> aic7860: Ultra Single Channel A, SCSI Id=7, 3/253 SCBs
(scsi3:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi generic sg5 at scsi3, channel 0, id 4, lun 0, type 1 st: Version 20040403, fixed bufsize 32768, s/g segs 256 Attached scsi tape st0 at scsi3, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
I did check termination. I used to use SCSI a lot, but it's been a few year. The SCSI card has auto termination built-in at it's end and I had the jumpers installed on the tape drive for Term Power and Term Act to be turned on. I also tried this new cable with a big terminator pack attached to the end of the SCSI cable with the tape drive connected at the connector just before it.
Any other ideas???
Chris Ruprecht wrote:
When the SCSI BIOS shows up on your boot screen, does it see the CD-ROM drive? If not, the most likely cause for it not beeing seen is that it is not working. Usually, you do not need a driver for SCSI devices - I have never needed one in my 11 years of Linux. Things to check: Does the CD-ROM drive have a unique SCSI-ID? This is settable with a few (3) jumpers and is binary encoded. No jumper -> SCSI-ID 0, 3 jumpers -> SCSI ID 7 - which is usually used by the Adaptor, so chose 0 - 6 only.
Do you have the SCSI termination right? The last device on the kable needs to be terminated or you have to add a terminator (not Arnold) to the last connector of the SCSI cable. You should also set the adaptor to be self terminating and make it the first/last device on the cable. Never have the adaptor sitting somewhere in the middle.
Best regards, Chris
On Mon, 2004-09-06 at 22:42, Gregory Gulik wrote:
I did check termination. I used to use SCSI a lot, but it's been a few year. The SCSI card has auto termination built-in at it's end and I had the jumpers installed on the tape drive for Term Power and Term Act to be turned on. I also tried this new cable with a big terminator pack attached to the end of the SCSI cable with the tape drive connected at the connector just before it.
You didn't say, but you did disable termination in the tape drive when you tried the cable that has its own terminator, right?
Any other ideas???
I would ask your brother-in-law to try your tape drive in his system. If it works, get him to try your SCSI adapter. Unless you know for sure that it works, the test you did with the old CD-ROM drive you have only proves there is a problem, which we know.
On Monday 06 September 2004 19:42, Gregory Gulik wrote:
Sorry for the delay in getting back to everyone on this.
Anyway, quick update, my SCSI tape drive doesn't work. The original post was here:
http://www.redhat.com/archives/fedora-list/2004-August/msg05266.html
First, to answer the question below, yes, the CD-ROM drive and the tape drive are found by the SCSI card BIOS screen just fine. If I use the built-in utilities the device comes up just fine as well.
Anyway, I tried several things and something isn't right. My brother-in-law gave me his known working Adaptec 2940U card, same model as mine, just appears to be a newer revision. I also have a different internal SCSI cable from him, this one with a terminator built-in to one end.
After rebooting Kudzu found something changed and had me unconfigure the old card and configure the new card. Fine with me.
However when it finished booting I found the card wasn't really recognized. The aic7xxxx driver had not been loaded. WTF!
If I load it manually with "modprobe aic7xxx" it shows up in dmesg:
scsi3 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940A Ultra SCSI adapter> aic7860: Ultra Single Channel A, SCSI Id=7, 3/253 SCBs
(scsi3:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi generic sg5 at scsi3, channel 0, id 4, lun 0, type 1 st: Version 20040403, fixed bufsize 32768, s/g segs 256 Attached scsi tape st0 at scsi3, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
I did check termination. I used to use SCSI a lot, but it's been a few year. The SCSI card has auto termination built-in at it's end and I had the jumpers installed on the tape drive for Term Power and Term Act to be turned on. I also tried this new cable with a big terminator pack attached to the end of the SCSI cable with the tape drive connected at the connector just before it.
***************************************** If you have a terminator on the end of the cable then the on-drive terminator should be set to OFF or NOT ACTIVE. Having two terminators causes an improper load on the controller resulting in errors. Don't think that is the REAL problem here though.
It's been several years since I used a 2940 and don't remember the cards configuration settings. Have you tried the Adaptec and Seagate (they bought Conner) web sites?
Tom *****************************************
Any other ideas???
Chris Ruprecht wrote:
When the SCSI BIOS shows up on your boot screen, does it see the CD-ROM drive? If not, the most likely cause for it not beeing seen is that it is not working. Usually, you do not need a driver for SCSI devices - I have never needed one in my 11 years of Linux. Things to check: Does the CD-ROM drive have a unique SCSI-ID? This is settable with a few (3) jumpers and is binary encoded. No jumper -> SCSI-ID 0, 3 jumpers -> SCSI ID 7 - which is usually used by the Adaptor, so chose 0 - 6 only.
Do you have the SCSI termination right? The last device on the kable needs to be terminated or you have to add a terminator (not Arnold) to the last connector of the SCSI cable. You should also set the adaptor to be self terminating and make it the first/last device on the cable. Never have the adaptor sitting somewhere in the middle.
Best regards, Chris
-- Greg Gulik http://www.gulik.org/greg/ greg @ gulik.org
Tom Taylor wrote:
If you have a terminator on the end of the cable then the on-drive terminator should be set to OFF or NOT ACTIVE. Having two terminators causes an improper load on the controller resulting in errors. Don't think that is the REAL problem here though.
Yes, when I switched to the cable with the built in terminator I disabled the tape drive's on-board termination.
It's been several years since I used a 2940 and don't remember the cards configuration settings. Have you tried the Adaptec and Seagate (they bought Conner) web sites?
The Adaptec 2940s are fairly basic SCSI cards that have been around for ages. I used to use this card and tape drive in a different box several years ago with Red Hat 7.2 and 7.3 and never had any problems with it. "It just worked" back then.
C. Linus Hicks wrote:
You didn't say, but you did disable termination in the tape drive when you tried the cable that has its own terminator, right?
Yes, I disabled the tape drive's on-board termination.
I would ask your brother-in-law to try your tape drive in his system. If it works, get him to try your SCSI adapter. Unless you know for sure that it works, the test you did with the old CD-ROM drive you have only proves there is a problem, which we know.
I'll have to do that next.
I did try the same card and tape drive combo in an older piece of hardware I have, which is also running Fedora Core 2. I had the exact same issues there.
To summarize the symptoms:
The card's BIOS recognizes the drive. Every time the bus gets reset the tape in the drive re-tensions.
Upon boot with the card kudzu asks me if I want to configure the card. I tell it to go ahead and configure it.
When done booting I find that the device is not seen and the aic7xxx driver is NOT loaded.
When I manually load the aix7xxx driver with modprobe the tape re-tensions and dmesg shows the following: scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.36 <Adaptec 2940A Ultra SCSI adapter> aic7860: Ultra Single Channel A, SCSI Id=7, 3/253 SCBs
(scsi1:A:4): 5.000MB/s transfers (5.000MHz, offset 15) Vendor: CONNER Model: CTT8000-S Rev: 1.17 Type: Sequential-Access ANSI SCSI revision: 02 Attached scsi generic sg1 at scsi1, channel 0, id 4, lun 0, type 1 st: Version 20040403, fixed bufsize 32768, s/g segs 256 Attached scsi tape st0 at scsi1, channel 0, id 4, lun 0 st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
Any attempt to access the st0 tape device results in the following:
[root@crashnburn root]# mt -f /dev/st0 status /dev/st0: Input/output error [root@crashnburn root]# gtar tzf /dev/st0 tar (child): /dev/st0: Cannot open: Input/output error tar (child): Error is not recoverable: exiting now
gzip: stdin: unexpected end of file gtar: Child returned status 2 gtar: Error exit delayed from previous errors
Do IDE tape drives work in Fedora? Will my backups be compatible with a different drive? This drive doesn't have hardware compression.
-
C. Linus Hicks -
Charles Curley -
Chris Ruprecht -
drcaaï¼ ig.com.br -
Gregory Gulik -
Jeff Vian -
Les Mikesell -
Michael Jezierski - Systec Conveyors -
Robert -
Thomas Taylor