|
Home > Archive > Unix administration > January 2004 > Problem communicating with AIT tape drive.
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Problem communicating with AIT tape drive.
|
|
| Peter Cattaneo 2004-01-23, 4:29 pm |
| I have an AIT drive that I can't communicate with anymore. It's been a few
weeks since I last used it. My Sun can't see the drive. I'm running
Solaris 2.6.
I haven't changed devices or anything. I haven't touched the machine
physically.
I did a <boot-r> to reconfigure the system. When I type <dmesg>, the drive
is not in the list.
I shutdown the machine and did a <probe-scsi-all> . I can see the drive.
It's on a chain with two other tape drives. Both of those drives work fine.
Can anyone give some advice on what I should try to do to get the drive up
and running again.
I assume it's a software OS problem since I can see the drive with
probe-scsi-all.
Any advice would be appreciated.
Peter
| |
| Michael Vilain 2004-01-23, 4:29 pm |
| In article <3f1d41d1.0@news.syr.edu>,
"Peter Cattaneo" <pkcattan@syr.edu> wrote:
quote:
> I have an AIT drive that I can't communicate with anymore. It's been a few
> weeks since I last used it. My Sun can't see the drive. I'm running
> Solaris 2.6.
> I haven't changed devices or anything. I haven't touched the machine
> physically.
> I did a <boot-r> to reconfigure the system. When I type <dmesg>, the drive
> is not in the list.
> I shutdown the machine and did a <probe-scsi-all> . I can see the drive.
> It's on a chain with two other tape drives. Both of those drives work fine.
>
> Can anyone give some advice on what I should try to do to get the drive up
> and running again.
>
> I assume it's a software OS problem since I can see the drive with
> probe-scsi-all.
>
> Any advice would be appreciated.
Can you see the other tape drives in the chain? What happens if you use
mt to get status on the device file?
You could clean out the /dev/rmt/ directory and do another boot -r. If
you still see the problem, replace the tape drive.
--
DeeDee, don't press that button! DeeDee! NO! Dee...
| |
| Michael Vilain 2004-01-23, 4:29 pm |
| In article <3f1d41d1.0@news.syr.edu>,
"Peter Cattaneo" <pkcattan@syr.edu> wrote:
quote:
> I have an AIT drive that I can't communicate with anymore. It's been a few
> weeks since I last used it. My Sun can't see the drive. I'm running
> Solaris 2.6.
> I haven't changed devices or anything. I haven't touched the machine
> physically.
> I did a <boot-r> to reconfigure the system. When I type <dmesg>, the drive
> is not in the list.
> I shutdown the machine and did a <probe-scsi-all> . I can see the drive.
> It's on a chain with two other tape drives. Both of those drives work fine.
>
> Can anyone give some advice on what I should try to do to get the drive up
> and running again.
>
> I assume it's a software OS problem since I can see the drive with
> probe-scsi-all.
>
> Any advice would be appreciated.
Can you see the other tape drives in the chain? What happens if you use
mt to get status on the device file?
You could clean out the /dev/rmt/ directory and do another boot -r. If
you still see the problem, replace the tape drive.
--
DeeDee, don't press that button! DeeDee! NO! Dee...
| |
| Michael Vilain 2004-01-23, 4:29 pm |
| In article <3f1e9bf6.0@news.syr.edu>,
"Peter Cattaneo" <pkcattan@syr.edu> wrote:
[QUOTE][color=darkred]
> No SCSI ID conflicts. The whole system hasn't been touched at all.
> It was working a few months ago. Again, system hasn't been touched since it
> was working.
>
>
> When I attempt mt commands, it says
> /dev/rmt/2: No such device or address
>
> The drive powers up, and I can load a tape. I can even eject tapes by
> pushing the eject button.
>
> The two other drives on the chain are working fine. This AIT drive is the
> first one on the chain.
>
> Any other suggestions? Should I remove the 2* files from /dev/rmt and do a
> boot -r?
>
> I thinking now the drive is bad...
>
> "Anthony Mandic" <mc@hotmail.com> wrote in message
> news:3F1E11D8.D4DE837D@hotmail.com...
[pay no attention to that little man behind the curtain from down under.
he's already in my killfile]
Since the OP said that nothing's changed on the system, I took them at
their word. I figured that something that's been working then stops is
hardware most of the time. Happens with tape drives, especially when
they're used regularly for backups (although the OP said his wasn't).
If "boot -r" won't scan and recreate the device file, there's a problem,
yes? Wipe the device files for all mt devices and boot -r. If they all
show up, great. If the problem drive doesn't, replace it (it's on
contract, right?).
If no drives appear, could be a bad cable or terminator (but do they
break all of a sudden--never seen that). Remove the first one and try
boot -r again.
The solution tree to this could go on quite a while. I gave the most
obvious (to me) possibility. Tape drives go bad sometimes. That's why
having them on contract saves you bucks when they break.
--
DeeDee, don't press that button! DeeDee! NO! Dee...
| |
| Cypherpunk@nyc.rr.com 2004-01-23, 4:29 pm |
| In comp.unix.solaris "Michael wrote:quote:
> "Peter Cattaneo" <pkcattan@syr.edu> wrote:
>
I've never heard of having to remove /dev entries to get
a reconfiguration boot to work. If you're lucky, devfsadm
can do it while the system is up.
FYI, to initiate a reconfiguration reboot from root,
type 'reboot - -r'.
----
"Michael wrote:[QUOTE][color=darkred]
> "Peter Cattaneo" <pkcattan@syr.edu> wrote:
Actually, no he didn't. He only stated the probe-scsi-all let him
see the problem drive. He also said the other drives work, but that
doesn't mean there's anything wrong with asking if they were also
listed in the probe-scsi-all. After all, there's already the discrepancy
of seeing a drive but it not working, so why not ask about the reverse?
----
"Michael wrote:[QUOTE][color=darkred]
> What happens if you use mt to get status on the device file?
>
"Anythony Manic" <mc@hotmail.com> wrote:quote:
> How is that going to work?
Since the OP said the probe-scsi-all could see the drive,
it was a reasonable suggestion to make. The OP didn't post
about "device doesn't exist" until a _subsequent_ post.
----
"Michael wrote:quote:
> "Peter Cattaneo" <pkcattan@syr.edu> wrote:
>
> [pay no attention to that little man behind the curtain from down under.
> he's already in my killfile]
Anythony is the Official Pest of CUS.
Given his above responses, one wonders why he even tries
to answer questions here.
Anythony had no suggestions to make. He only posted to ridicule.
Hey, Anythony, care to make another sexual crack about me?
"Michael wrote:quote:
> Remove the first one and try boot -r again.
Power down everything.
Remove ALL other attached SCSI devices and do a reconfiguration boot.
See what's in /dev/rmt.
If that works, you know what to do from there...
----
Check /var/adm/messages* for AIT messages.
Don't some of those AIT drives have special software (device drivers?)
that do something funky to set the units to read-only to guarantee
a operator can't overwrite a backup tape?
Maybe it's something really obscure, like a script for the AIT's
in /etc/rc*.d/* that was set to delete the /dev entry if the
protective device driver isn't in the booted kernel.
Or is that a tape drive called a GRAU?
| |
| Michael Vilain 2004-01-23, 4:29 pm |
| In article <3f1e9bf6.0@news.syr.edu>,
"Peter Cattaneo" <pkcattan@syr.edu> wrote:
[QUOTE][color=darkred]
> No SCSI ID conflicts. The whole system hasn't been touched at all.
> It was working a few months ago. Again, system hasn't been touched since it
> was working.
>
>
> When I attempt mt commands, it says
> /dev/rmt/2: No such device or address
>
> The drive powers up, and I can load a tape. I can even eject tapes by
> pushing the eject button.
>
> The two other drives on the chain are working fine. This AIT drive is the
> first one on the chain.
>
> Any other suggestions? Should I remove the 2* files from /dev/rmt and do a
> boot -r?
>
> I thinking now the drive is bad...
>
> "Anthony Mandic" <mc@hotmail.com> wrote in message
> news:3F1E11D8.D4DE837D@hotmail.com...
[pay no attention to that little man behind the curtain from down under.
he's already in my killfile]
Since the OP said that nothing's changed on the system, I took them at
their word. I figured that something that's been working then stops is
hardware most of the time. Happens with tape drives, especially when
they're used regularly for backups (although the OP said his wasn't).
If "boot -r" won't scan and recreate the device file, there's a problem,
yes? Wipe the device files for all mt devices and boot -r. If they all
show up, great. If the problem drive doesn't, replace it (it's on
contract, right?).
If no drives appear, could be a bad cable or terminator (but do they
break all of a sudden--never seen that). Remove the first one and try
boot -r again.
The solution tree to this could go on quite a while. I gave the most
obvious (to me) possibility. Tape drives go bad sometimes. That's why
having them on contract saves you bucks when they break.
--
DeeDee, don't press that button! DeeDee! NO! Dee...
| |
| Cypherpunk@nyc.rr.com 2004-01-23, 4:29 pm |
| In comp.unix.solaris "Michael wrote:quote:
> "Peter Cattaneo" <pkcattan@syr.edu> wrote:
>
I've never heard of having to remove /dev entries to get
a reconfiguration boot to work. If you're lucky, devfsadm
can do it while the system is up.
FYI, to initiate a reconfiguration reboot from root,
type 'reboot - -r'.
----
"Michael wrote:[QUOTE][color=darkred]
> "Peter Cattaneo" <pkcattan@syr.edu> wrote:
Actually, no he didn't. He only stated the probe-scsi-all let him
see the problem drive. He also said the other drives work, but that
doesn't mean there's anything wrong with asking if they were also
listed in the probe-scsi-all. After all, there's already the discrepancy
of seeing a drive but it not working, so why not ask about the reverse?
----
"Michael wrote:[QUOTE][color=darkred]
> What happens if you use mt to get status on the device file?
>
"Anythony Manic" <mc@hotmail.com> wrote:quote:
> How is that going to work?
Since the OP said the probe-scsi-all could see the drive,
it was a reasonable suggestion to make. The OP didn't post
about "device doesn't exist" until a _subsequent_ post.
----
"Michael wrote:quote:
> "Peter Cattaneo" <pkcattan@syr.edu> wrote:
>
> [pay no attention to that little man behind the curtain from down under.
> he's already in my killfile]
Anythony is the Official Pest of CUS.
Given his above responses, one wonders why he even tries
to answer questions here.
Anythony had no suggestions to make. He only posted to ridicule.
Hey, Anythony, care to make another sexual crack about me?
"Michael wrote:quote:
> Remove the first one and try boot -r again.
Power down everything.
Remove ALL other attached SCSI devices and do a reconfiguration boot.
See what's in /dev/rmt.
If that works, you know what to do from there...
----
Check /var/adm/messages* for AIT messages.
Don't some of those AIT drives have special software (device drivers?)
that do something funky to set the units to read-only to guarantee
a operator can't overwrite a backup tape?
Maybe it's something really obscure, like a script for the AIT's
in /etc/rc*.d/* that was set to delete the /dev entry if the
protective device driver isn't in the booted kernel.
Or is that a tape drive called a GRAU?
|
|
|
|
|