Claude Jones claude_jones@levitjames.com wrote:
Host: scsi0 Channel: 00 Id: 00 Lun: 00 <--- Vendor: ATA Model: ST3160023AS Rev: 3.05 Type: Direct-Access ANSI SCSI revision: 05 Host: scsi1 Channel: 00 Id: 00 Lun: 00 <--- Vendor: ATA Model: ST3300622AS Rev: 3.AA Type: Direct-Access ANSI SCSI revision: 05
It looks like the SCSI ID being assigned to these drives is the same. The ordered triple (Channel, ID, LUN) should be unique for each device. Don't know how this can happen since it should never happen.
Unfortunately, all of the relics here don't use SATA drives so I can't take this any further. Also, it looks like the system sorts it all out by the time you get a command prompt so everything eventually works; you just can't do things automatically during the boot process.
Past my bedtime here. Hope this helps.
Cheers, Dave
On Thu November 2 2006 2:09 am, David G. Miller wrote:
Claude Jones claude_jones@levitjames.com wrote:
Host: scsi0 Channel: 00 Id: 00 Lun: 00 <--- Vendor: ATA Model: ST3160023AS Rev: 3.05 Type: Direct-Access ANSI SCSI revision: 05 Host: scsi1 Channel: 00 Id: 00 Lun: 00 <--- Vendor: ATA Model: ST3300622AS Rev: 3.AA Type: Direct-Access ANSI SCSI revision: 05
It looks like the SCSI ID being assigned to these drives is the same. The ordered triple (Channel, ID, LUN) should be unique for each device. Don't know how this can happen since it should never happen.
Unfortunately, all of the relics here don't use SATA drives so I can't take this any further. Also, it looks like the system sorts it all out by the time you get a command prompt so everything eventually works; you just can't do things automatically during the boot process.
Past my bedtime here. Hope this helps.
Sure it helps. I did see that issue, but didn't know enough to know whether that was abnormal. I can see where that could cause problems in the early boot process, if it is seeing two drives as the same drive, in effect... I'm going to start a new SATA thread on this, and see if any responds. Could just be I've found a SATA bug...
On Thu November 2 2006 7:39 am, Claude Jones wrote:
On Thu November 2 2006 2:09 am, David G. Miller wrote:
Claude Jones claude_jones@levitjames.com wrote:
Host: scsi0 Channel: 00 Id: 00 Lun: 00 <--- Vendor: ATA Model: ST3160023AS Rev: 3.05 Type: Direct-Access ANSI SCSI revision: 05 Host: scsi1 Channel: 00 Id: 00 Lun: 00 <--- Vendor: ATA Model: ST3300622AS Rev: 3.AA Type: Direct-Access ANSI SCSI revision: 05
It looks like the SCSI ID being assigned to these drives is the same. The ordered triple (Channel, ID, LUN) should be unique for each device. Don't know how this can happen since it should never happen.
Unfortunately, all of the relics here don't use SATA drives so I can't take this any further. Also, it looks like the system sorts it all out by the time you get a command prompt so everything eventually works; you just can't do things automatically during the boot process.
Past my bedtime here. Hope this helps.
Sure it helps. I did see that issue, but didn't know enough to know whether that was abnormal. I can see where that could cause problems in the early boot process, if it is seeing two drives as the same drive, in effect... I'm going to start a new SATA thread on this, and see if any responds. Could just be I've found a SATA bug...
For the record: Paul Howarth had suggested a fix which post I somehow missed, to wit:
"With the drive not mounted, do: $ chcon -t mnt_t /home/cj/archive
You can then simulate the attempted mount at boot time: # service netfs start
Alternatively you could just reboot to check that the problem is fixed for sure.
Paul."
This has fixed the problem...
On Thu, 2006-11-02 at 10:14 -0500, Claude Jones wrote:
For the record: Paul Howarth had suggested a fix which post I somehow missed, to wit:
"With the drive not mounted, do: $ chcon -t mnt_t /home/cj/archive
You can then simulate the attempted mount at boot time: # service netfs start
Wouldn't that (the service netfs start) only affect mounts that traverse a network?
On Thu November 2 2006 10:39 am, Tim wrote:
Wouldn't that (the service netfs start) only affect mounts that traverse a network?
Don't know. It worked, though. And, upon reboot, the drive was mounted normally.
Tim wrote:
On Thu, 2006-11-02 at 10:14 -0500, Claude Jones wrote:
For the record: Paul Howarth had suggested a fix which post I somehow missed, to wit:
"With the drive not mounted, do: $ chcon -t mnt_t /home/cj/archive
You can then simulate the attempted mount at boot time: # service netfs start
Wouldn't that (the service netfs start) only affect mounts that traverse a network?
That's what you might think given the name, but it actually tries to mount everything, network-based or not. Try unmounting a local filesystem you have listed in fstab and see for yourself.
Paul.