Dear All
I have added the following line to my /etc/vfstab :
/dev/dsk/c0t2d0s3 /dev/rdsk/c0t2d0s3 /opt ufs 2 yes -
Then I made an image from another solaris8 machine /opt and added it
to mine. At now, the format command recognizes three disks but my /opt
does not contain any data. Can you please let me know why?

On Tue, 2010-07-06 at 20:58 -0700, hadi motamedi wrote:
> Dear All
> I have added the following line to my /etc/vfstab :
> /dev/dsk/c0t2d0s3 /dev/rdsk/c0t2d0s3 /opt ufs 2 yes -
> Then I made an image from another solaris8 machine /opt and added it
> to mine. At now, the format command recognizes three disks but my /opt
> does not contain any data. Can you please let me know why?

Maybe /opt is a part of the root filesystem data (/) and didn't really
get made into it's own filesystem??  Just guessing...

> Maybe /opt is a part of the root filesystem data (/) and didn't really
> get made into it's own filesystem?? =A0Just guessing...
The '/' resides on c0t0d0s0 but the '/opt' resides on /c0t0d0s5 so on
different slices . Can you please let me know how to add data from the
imported /opt ?

On Jul 7, 2:08=A0pm, hadi motamedi <motamed...@gmail.com> wrote:
> > Maybe /opt is a part of the root filesystem data (/) and didn't really
> > get made into it's own filesystem?? =A0Just guessing...
>
> The '/' resides on c0t0d0s0 but the '/opt' resides on /c0t0d0s5 so on
> different slices . Can you please let me know how to add data from the
> imported /opt ?

{assuming that your new /opt is mount properly}

mount /dev/dsk/c0t0d0s5 /mnt
cd /opt
ufsdump 0f /mnt - | ufsrestore -rf -
umount /mnt

That will copy what was in your old /opt to your new /opt on the new
disk.

> mount /dev/dsk/c0t0d0s5 /mnt
> cd /opt
> ufsdump 0f /mnt - | ufsrestore -rf -
> umount /mnt
> That will copy what was in your old /opt to your new /opt on the new
> disk.
My old(original) /opt resides on /c0t0d0s5 and it has no data and I
want to import new data from added /c0t2d0s3 to /opt. Can you please
let me know how it can be accomplished?

On Jul 7, 2:57=A0pm, hadi motamedi <motamed...@gmail.com> wrote:
> > mount /dev/dsk/c0t0d0s5 /mnt
> > cd /opt
> > ufsdump 0f /mnt - | ufsrestore -rf -
> > umount /mnt
> > That will copy what was in your old /opt to your new /opt on the new
> > disk.
>
> My old(original) /opt resides on /c0t0d0s5 and it has no data and I
> want to import new data from added /c0t2d0s3 to /opt. Can you please
> let me know how it can be accomplished?

Procedure would be pretty much the same then.  Just change what you
are mounting.

mount /dev/dsk/c0t2d0s3 /mnt
cd /opt
ufsdump 0f - /mnt | ufsrestore -rf -
umount /mnt

Please not that I screwed up the arguments for the ufsdump first time
around.
The name of the file system to dump should come after the file to dump
to ('-' i.e.
stdout.)  Should have check the man page first.

On Tue, 06 Jul 2010 22:08:51 -0700, hadi motamedi wrote:

>> Maybe /opt is a part of the root filesystem data (/) and didn't really
>> get made into it's own filesystem?? �Just guessing...
> The '/' resides on c0t0d0s0 but the '/opt' resides on /c0t0d0s5 so on
> different slices . Can you please let me know how to add data from the
> imported /opt ?

I'd be interested in why /opt is empty ?

If you want to add a list of packages cleanly you should really get the
package(s) and use pkgadd.

Otherwise future admin actions, like patching, may not turn out very well.


-- 
	Bruce
"The internet is a huge and diverse community and not every one is friendly" 		           
        http://www.ytc1.co.uk                     

hadi motamedi <motamedi24@gmail.com> wrote:
> My old(original) /opt resides on /c0t0d0s5 and it has no data and I
> want to import new data from added /c0t2d0s3 to /opt. Can you please
> let me know how it can be accomplished?

Are you mounting your new /opt first, and then mounting the old, empty /opt
on top of it?  Did you make sure to remove or comment out the old /opt from
your /etc/vfstab?  When you do a 'df -k /opt', what device does it show?

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On Tue, 6 Jul 2010 22:08:51 -0700 (PDT), hadi motamedi
<motamedi24@gmail.com> wrote:

>> Maybe /opt is a part of the root filesystem data (/) and didn't really
>> get made into it's own filesystem?? �Just guessing...
>The '/' resides on c0t0d0s0 but the '/opt' resides on /c0t0d0s5 so on
>different slices . Can you please let me know how to add data from the
>imported /opt ?

Can you copy it?

> Are you mounting your new /opt first, and then mounting the old, empty /o=
pt
> on top of it? =A0Did you make sure to remove or comment out the old /opt =
from
> your /etc/vfstab? =A0When you do a 'df -k /opt', what device does it show=
?
Thank you for your reply. I commented out the original /opt from my /
etc/vfstab and then added entry for mounting new slice on it. But
still it does not contain new imported data and the 'df -k' shows it
mounted on /dev/dsk/c0t0d0s0 . Can you please let me know how to
import new data inside it?

hadi motamedi <motamedi24@gmail.com> wrote:
> mounted on /dev/dsk/c0t0d0s0 . Can you please let me know how to
> import new data inside it?

c0t0d0s0 is likely your root slice.  So your /opt isn't being mounted at all,
I'd bet.

What happens if you try the command "mount /opt"?

What does "grep opt /etc/vfstab" tell you?

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

> What happens if you try the command "mount /opt"?
> What does "grep opt /etc/vfstab" tell you?
The original /opt resides on /c0t0d0s5 , so I added a line to my /etc/
vfstab to put it on the new imported /c0t2d0s3 . The 'mount /opt'
command will mount it on the /c0t0d0s5 and the output of 'grep opt /
etc/vfstab' is as the following:
/dev/dsk/c0t0d0s5 /dev/rdsk/c0t0d0s5 /opt ufs 2 yes -
/dev/dsk/c0t2d0s3 /dev/rdsk/c0t2d0s3 /opt ufs 2 yes -

On 07/12/10 03:42 PM, hadi motamedi wrote:

[please stop stripping attributions, it's rude]

>> What happens if you try the command "mount /opt"?
>> What does "grep opt /etc/vfstab" tell you?
> The original /opt resides on /c0t0d0s5 , so I added a line to my /etc/
> vfstab to put it on the new imported /c0t2d0s3 . The 'mount /opt'
> command will mount it on the /c0t0d0s5 and the output of 'grep opt /
> etc/vfstab' is as the following:
> /dev/dsk/c0t0d0s5 /dev/rdsk/c0t0d0s5 /opt ufs 2 yes -
> /dev/dsk/c0t2d0s3 /dev/rdsk/c0t2d0s3 /opt ufs 2 yes -

You can't have two devices on the same mountpoint!

Remove the first line and manually mount /dev/dsk/c0t0d0s5 somewhere 
else (say /mnt) and copy the data over.

-- 
Ian Collins

> Remove the first line and manually mount /dev/dsk/c0t0d0s5 somewhere
> else (say /mnt) and copy the data over.
I removed the first line from /etc/vfstab and I tried to manually
mount /dev/dsk/c0t0d0s5 on /mnt but it is returning '/mnt is busy' .
Can you please let me know how can I implement your proposed
procedure ?

hadi motamedi <motamedi24@gmail.com> wrote:
> I removed the first line from /etc/vfstab and I tried to manually
> mount /dev/dsk/c0t0d0s5 on /mnt but it is returning '/mnt is busy' .

Are you standing in /mnt when you try to mount that filesystems?  What
I mean is, is your current working directory /mnt at the time?

As an example (using ZFS, when you are using UFS, but that doesn't matter):

  shumira:/export/home/hume# cd /mnt
  shumira:/mnt# mount -F zfs rpool/test /mnt
  mount failed: Device busy
  shumira:/mnt# cd /
  shumira:/# mount -F zfs rpool/test /mnt
  shumira:/# 

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On 07/12/10 07:33 PM, hadi motamedi wrote:
>> Remove the first line and manually mount /dev/dsk/c0t0d0s5 somewhere
>> else (say /mnt) and copy the data over.
> I removed the first line from /etc/vfstab and I tried to manually
> mount /dev/dsk/c0t0d0s5 on /mnt but it is returning '/mnt is busy' .
> Can you please let me know how can I implement your proposed
> procedure ?

Maybe if you stop snipping attributions!

-- 
Ian Collins

On Jul 13, 6:28=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> On 07/12/10 07:33 PM, hadi motamedi wrote:
>
> >> Remove the first line and manually mount /dev/dsk/c0t0d0s5 somewhere
> >> else (say /mnt) and copy the data over.
> > I removed the first line from /etc/vfstab and I tried to manually
> > mount /dev/dsk/c0t0d0s5 on /mnt but it is returning '/mnt is busy' .
> > Can you please let me know how can I implement your proposed
> > procedure ?
>
> Maybe if you stop snipping attributions!
>
> --
> Ian Collins

It also might help if he were to go back and re-explain what sort of
state the
box is in.  He has been given quite a lot of advice and I get the
feeling that
he has applied bits of it here and there trying to get things to work
and now
the box is likely in a half-assed stated.

> Are you standing in /mnt when you try to mount that filesystems? =A0What
> I mean is, is your current working directory /mnt at the time?
> As an example (using ZFS, when you are using UFS, but that doesn't matter=
):
> =A0 shumira:/export/home/hume# cd /mnt
> =A0 shumira:/mnt# mount -F zfs rpool/test /mnt
> =A0 mount failed: Device busy
> =A0 shumira:/mnt# cd /
> =A0 shumira:/# mount -F zfs rpool/test /mnt
> =A0 shumira:/#
> --
> Brandon Hume =A0 =A0- hume -> BOFH.Ca,http://WWW.BOFH.Ca/
My imported data resides in /dev/dsk/c0t2d0s3 that I want to put it
in /opt . According to you, I removed my original /opt definition
from /etc/vfstab . I mounted the original /opt slice (/dev/dsk/
c0t0d0s5) on /mnt but when I want to manually mount the new data on /
opt, I got the following error :
# cd /
# mount -F ufs /dev/dsk/c0t0d0s5 /mnt
# mount -F ufs /dev/dsk/c0t2d0s3 /opt
WARNING: /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/
esp@5,8800000/sd@2,0
 (sd2):
        corrupt label - wrong magic number

mount: I/O error
mount: cannot mount /dev/dsk/c0t2d0s3
#
Please help me.
Thank you

On 07/12/10 11:50 PM, hadi motamedi wrote:
>>      
> My imported data resides in /dev/dsk/c0t2d0s3 that I want to put it
> in /opt . According to you, I removed my original /opt definition
> from /etc/vfstab . I mounted the original /opt slice (/dev/dsk/
> c0t0d0s5) on /mnt but when I want to manually mount the new data on /
> opt, I got the following error :
> # cd /
> # mount -F ufs /dev/dsk/c0t0d0s5 /mnt
> # mount -F ufs /dev/dsk/c0t2d0s3 /opt
> WARNING: /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/
> esp@5,8800000/sd@2,0
>   (sd2):
>          corrupt label - wrong magic number
>
> mount: I/O error
> mount: cannot mount /dev/dsk/c0t2d0s3
> #
>    
If I remember correctly (been on Opensolaris for 2 years now), "wrong 
magic number" usually means the disk is not formatted or doesn't have a 
proper label.  Have you run the "format" command and then "mkfs" command 
or "newfs" command on that disk?  Please note, "mkfs" and "newfs" have 
the same end result, "newfs" is easier to run.

Paul

>
> If I remember correctly (been on Opensolaris for 2 years now), "wrong
> magic number" usually means the disk is not formatted or doesn't have a
> proper label. =A0Have you run the "format" command and then "mkfs" comman=
d
> or "newfs" command on that disk? =A0Please note, "mkfs" and "newfs" have
> the same end result, "newfs" is easier to run.
>
> Paul

The format command shows three disks available :
'# format
Searching for disks...WARNING: /iommu@0,10000000/sbus@0,10001000/
espdma@5,840000
0/esp@5,8800000/sd@2,0 (sd2):
        corrupt label - wrong magic number
done
AVAILABLE DISK SELECTIONS:
       0. c0t0d0 <Qemu2GB cyl 4090 alt 2 hd 16 sec 63>
          /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/
esp@5,8800000/sd@0,
0
       1. c0t1d0 <Qemu2GB cyl 4090 alt 2 hd 16 sec 63>
          /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/
esp@5,8800000/sd@1,
0
       2. c0t2d0 <drive type unknown>
          /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/
esp@5,8800000/sd@2,
0'
Then I tried for 'newfs' on the third disk , as :
'# newfs /dev/dsk/c0t2d0s3
WARNING: /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/
esp@5,8800000/sd@2,0
 (sd2):
        corrupt label - wrong magic number
/dev/rdsk/c0t2d0s3: I/O error'
Can you please let me know how to correct it?
Thank you

>
> If I remember correctly (been on Opensolaris for 2 years now), "wrong
> magic number" usually means the disk is not formatted or doesn't have a
> proper label. =A0Have you run the "format" command and then "mkfs" comman=
d
> or "newfs" command on that disk? =A0Please note, "mkfs" and "newfs" have
> the same end result, "newfs" is easier to run.
>
> Paul

Sorry. With your help, it is now returning as :
'#mount /dev/dsk/c0t2d0s3 /opt
mount /dev/dsk/c0t2d0s3 is not this fstype'
Can you please let me know how to correct this?
Thank you

On Jul 13, 9:52=A0am, hadi motamedi <motamed...@gmail.com> wrote:
> > If I remember correctly (been on Opensolaris for 2 years now), "wrong
> > magic number" usually means the disk is not formatted or doesn't have a
> > proper label. =A0Have you run the "format" command and then "mkfs" comm=
and
> > or "newfs" command on that disk? =A0Please note, "mkfs" and "newfs" hav=
e
> > the same end result, "newfs" is easier to run.
>
> > Paul
>
> Sorry. With your help, it is now returning as :
> '#mount /dev/dsk/c0t2d0s3 /opt
> mount /dev/dsk/c0t2d0s3 is not this fstype'
> Can you please let me know how to correct this?
> Thank you

#mount -F ufs /dev/dsk/c0t2d0s3 /opt - in case you've created ufs file
system there.

>
> #mount -F ufs /dev/dsk/c0t2d0s3 /opt - in case you've created ufs file
> system there.
I tried to mount it as :
'# mount -F ufs /dev/dsk/c0t2d0s3 /opt
mount: /dev/dsk/c0t2d0s3 is already mounted, /opt is busy,
        or the allowable number of mount points has been exceeded
# ls -la /opt
total 20
drwxr-xr-x   3 root     root         512 Jul 13 16:10 .
drwxr-xr-x  21 root     root         512 Jul  7 13:42 ..
drwx------   2 root     root        8192 Jul 13 16:10 lost+found
# df -k
Filesystem            kbytes    used   avail capacity  Mounted on
/dev/dsk/c0t0d0s0      61615   14892   40562    27%    /
/dev/dsk/c0t0d0s6     384847   21080  325283     7%    /usr
/proc                      0       0       0     0%    /proc
fd                         0       0       0     0%    /dev/fd
/dev/dsk/c0t0d0s1     114143    1321  101408     2%    /var
/dev/dsk/c0t0d0s7     192423       9  173172     1%    /export/home
swap                  227444       4  227440     1%    /tmp
/dev/dsk/c0t1d0s2    1936412   10578 1732193     1%    /usr/local
/dev/dsk/c0t2d0s3     123231       9  123222     1%    /opt'
As you see, it is mounted but it does not contain the imported data.
Can you please let me know why ?

hadi motamedi <motamedi24@gmail.com> writes:

>>
>> #mount -F ufs /dev/dsk/c0t2d0s3 /opt - in case you've created ufs file
>> system there.
> I tried to mount it as :
> '# mount -F ufs /dev/dsk/c0t2d0s3 /opt
> mount: /dev/dsk/c0t2d0s3 is already mounted, /opt is busy,
>         or the allowable number of mount points has been exceeded
> # ls -la /opt
> total 20
> drwxr-xr-x   3 root     root         512 Jul 13 16:10 .
> drwxr-xr-x  21 root     root         512 Jul  7 13:42 ..
> drwx------   2 root     root        8192 Jul 13 16:10 lost+found
> # df -k
> Filesystem            kbytes    used   avail capacity  Mounted on
> /dev/dsk/c0t0d0s0      61615   14892   40562    27%    /
> /dev/dsk/c0t0d0s6     384847   21080  325283     7%    /usr
> /proc                      0       0       0     0%    /proc
> fd                         0       0       0     0%    /dev/fd
> /dev/dsk/c0t0d0s1     114143    1321  101408     2%    /var
> /dev/dsk/c0t0d0s7     192423       9  173172     1%    /export/home
> swap                  227444       4  227440     1%    /tmp
> /dev/dsk/c0t1d0s2    1936412   10578 1732193     1%    /usr/local
> /dev/dsk/c0t2d0s3     123231       9  123222     1%    /opt'
> As you see, it is mounted but it does not contain the imported data.
> Can you please let me know why ?

If you did newfs on it, you have created a new filesystem on that disk
and all old data is lost forever.

>
> If you did newfs on it, you have created a new filesystem on that disk
> and all old data is lost forever.

oh, yes I did it like the following :
#newfs -m 0 /dev/dsk/c0t2d0s3
I have a backup from it. Can you please let me know what I am expected
to do to preserve its data ?

On 2010-07-13 10:56:09 +0100, hadi motamedi said:

>> 
>> 
>> If you did newfs on it, you have created a new filesystem on that disk
>> and all old data is lost forever.
> 
> oh, yes I did it like the following :
> #newfs -m 0 /dev/dsk/c0t2d0s3
> I have a backup from it. Can you please let me know what I am expected
> to do to preserve its data ?

Might it be simplest to restore the contents? For example if you took 
the backup with ufsdump, then use ufsrestore.
-- 
Chris

hadi motamedi <motamedi24@gmail.com> writes:

>>
>> If you did newfs on it, you have created a new filesystem on that disk
>> and all old data is lost forever.
>
> oh, yes I did it like the following :
> #newfs -m 0 /dev/dsk/c0t2d0s3
> I have a backup from it. Can you please let me know what I am expected
> to do to preserve its data ?

First, redo newfs without the -m option. -m 0 gives you 0% free space...

Then, to preserve is too late, you have to restore the data from your backup. 

# cd /mnt
# ufsrestore xvf /dev/rmt/0hn

On 07/13/10 09:56 PM, hadi motamedi wrote:

Who posted this?

>>
>> If you did newfs on it, you have created a new filesystem on that disk
>> and all old data is lost forever.
>
> oh, yes I did it like the following :
> #newfs -m 0 /dev/dsk/c0t2d0s3
> I have a backup from it. Can you please let me know what I am expected
> to do to preserve its data ?

Is this for real, or a troll?

-- 
Ian Collins

Ian Collins <ian-news@hotmail.com> writes:

> On 07/13/10 09:56 PM, hadi motamedi wrote:
>
> Who posted this?

OP I suppose.

>>>
>>> If you did newfs on it, you have created a new filesystem on that disk
>>> and all old data is lost forever.
>>
>> oh, yes I did it like the following :
>> #newfs -m 0 /dev/dsk/c0t2d0s3
>> I have a backup from it. Can you please let me know what I am expected
>> to do to preserve its data ?
>
> Is this for real, or a troll?

I'm afraid it's for real...

hadi motamedi <motamedi24@gmail.com> wrote:
> My imported data resides in /dev/dsk/c0t2d0s3 that I want to put it
> in /opt . According to you, I removed my original /opt definition

You say you have data in /dev/dsk/c0t2d0s3 that you want to copy into your
new /opt filesystem.  But then:

> # mount -F ufs /dev/dsk/c0t2d0s3 /opt
>        corrupt label - wrong magic number

This means that /dev/dsk/c0t2d0s3 is not a valid filesystem.  So it isn't
possible that you have data on it.  Obviously something you're telling us
is wrong.

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On Jul 13, 5:41=A0am, hume.spamfil...@bofh.ca wrote:
>
> You say you have data in /dev/dsk/c0t2d0s3 that you want to copy into you=
r
> new /opt filesystem. =A0But then:
>
> This means that /dev/dsk/c0t2d0s3 is not a valid filesystem. =A0So it isn=
't
> possible that you have data on it. =A0Obviously something you're telling =
us
> is wrong.
>
> --
> Brandon Hume =A0 =A0- hume -> BOFH.Ca,http://WWW.BOFH.Ca/

Please be informed that I have made image from my working solaris8 /
opt as the following :
#dd if=3D/dev/dsk/c0t0d0s3 bs=3D512 count=3D1966080 |gzip -c >/tmp/opt8.dis=
k-
pt1.gz
#dd if=3D/dev/dsk/c0t0d0s3 bs=3D512 count=3D1966080 skip=3D1966080 |gzip -c=
 >/
tmp/opt8.disk-pt2.gz
#dd if=3D/dev/dsk/c0t0d0s3 bs=3D512 count=3D1966080 skip=3D3932160 |gzip -c=
 >/
tmp/opt8.disk-pt3.gz
#dd if=3D/dev/dsk/c0t0d0s3 bs=3D512 count=3D1966080 skip=3D5898240 |gzip -c=
 >/
tmp/opt8.disk-pt4.gz
And so on .
At the end, I uncompressed all of them and concatenate as :
C:\copy /b opt8.disk-pt1+opt8.disk-pt2+opt8.disk-pt3+..... opt8.disk
Then I imported this 'opt8.disk' to my qemu-system-sparc running
solaris26 with the '-hdc opt8.disk' option . On my qemu solaris26, the
format command shows three disks attached and the 'df -k' command
shows the /opt is being mounted on /c0t2d0s3 (new imported slice) . Do
you think my procedure for creating the image from the solaris8
machine /opt is wrong that it did not lead to a slice with valid data
on it ? Can you please show me the correct procedure ?

hadi motamedi <motamedi24@gmail.com> wrote:
> you think my procedure for creating the image from the solaris8
> machine /opt is wrong that it did not lead to a slice with valid data
> on it ? Can you please show me the correct procedure ?

There is no correct procedure, because nobody has ever done what you're
trying to do.  People are guessing based on theory.

I can't think of a blatant reason why it wouldn't work, but I'd need to
try it myself to be sure.

I'm sorry to be so blunt, but your utter lack of Solaris skills combined
with the absolutely insane procedure you're attempting means you're very,
very unlikely to succeed.  You've spent how many months trying to do this?
And for some reason this is better than simply obtaining another Sun SPARC
machine?

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

>
> There is no correct procedure, because nobody has ever done what you're
> trying to do. =A0People are guessing based on theory.
>
> I can't think of a blatant reason why it wouldn't work, but I'd need to
> try it myself to be sure.
>
> I'm sorry to be so blunt, but your utter lack of Solaris skills combined
> with the absolutely insane procedure you're attempting means you're very,
> very unlikely to succeed. =A0You've spent how many months trying to do th=
is?
> And for some reason this is better than simply obtaining another Sun SPAR=
C
> machine?
>
> --
> Brandon Hume =A0 =A0- hume -> BOFH.Ca,http://WWW.BOFH.Ca/

At last, I sorted to do it by ufsdump/restore . On my working solaris8
machine :
#ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
I imported this as virtual disk three on my qemu with '-hdc opt8.disk'
option and its format command shows /c0t2d0 added as third disk. To
restore :
#ufsrestore -xvf /dev/dsk/c0t2d0s3
But it is returning as 'volume is not in dump format' . Can you please
let me know what is wrong here?

On 07/17/10 07:17 PM, hadi motamedi wrote:
>>
>> There is no correct procedure, because nobody has ever done what you're
>> trying to do.  People are guessing based on theory.
>>
>> I can't think of a blatant reason why it wouldn't work, but I'd need to
>> try it myself to be sure.
>>
>> I'm sorry to be so blunt, but your utter lack of Solaris skills combined
>> with the absolutely insane procedure you're attempting means you're very,
>> very unlikely to succeed.  You've spent how many months trying to do this?
>> And for some reason this is better than simply obtaining another Sun SPARC
>> machine?
>>
>> --
>> Brandon Hume    - hume ->  BOFH.Ca,http://WWW.BOFH.Ca/
>
> At last, I sorted to do it by ufsdump/restore . On my working solaris8
> machine :
> #ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
> I imported this as virtual disk three on my qemu with '-hdc opt8.disk'
> option and its format command shows /c0t2d0 added as third disk. To
> restore :
> #ufsrestore -xvf /dev/dsk/c0t2d0s3
> But it is returning as 'volume is not in dump format' . Can you please
> let me know what is wrong here?

Only if you stop snipping attributions!

-- 
Ian Collins

ok , I understand. Please let me know why my ufsrestore does not work
here ?

On Jul 17, 12:22=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> On 07/17/10 07:17 PM, hadi motamedi wrote:
>
>
>
>
>
> >> There is no correct procedure, because nobody has ever done what you'r=
e
> >> trying to do. =A0People are guessing based on theory.
>
> >> I can't think of a blatant reason why it wouldn't work, but I'd need t=
o
> >> try it myself to be sure.
>
> >> I'm sorry to be so blunt, but your utter lack of Solaris skills combin=
ed
> >> with the absolutely insane procedure you're attempting means you're ve=
ry,
> >> very unlikely to succeed. =A0You've spent how many months trying to do=
 this?
> >> And for some reason this is better than simply obtaining another Sun S=
PARC
> >> machine?
>
> >> --
> >> Brandon Hume =A0 =A0- hume -> =A0BOFH.Ca,http://WWW.BOFH.Ca/
>
> > At last, I sorted to do it by ufsdump/restore . On my working solaris8
> > machine :
> > #ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
> > I imported this as virtual disk three on my qemu with '-hdc opt8.disk'
> > option and its format command shows /c0t2d0 added as third disk. To
> > restore :
> > #ufsrestore -xvf /dev/dsk/c0t2d0s3
> > But it is returning as 'volume is not in dump format' . Can you please
> > let me know what is wrong here?
>
> Only if you stop snipping attributions!
>
> --
> Ian Collins

Sorry. What do you mean by 'stop snipping attributions' ?

On 2010-07-17 09:45:29 +0100, hadi motamedi said:

> On Jul 17, 12:22�am, Ian Collins <ian-n...@hotmail.com> wrote:
>> On 07/17/10 07:17 PM, hadi motamedi wrote:
>>> let me know what is wrong here?
>> 
>> Only if you stop snipping attributions!
>> 
>> --
>> Ian Collins
> 
> Sorry. What do you mean by 'stop snipping attributions' ?

You tend to leave out the lines I've quoted above (eg On 2010-07-17...) 
which let people see who wrote what.
-- 
Chris

On Jul 17, 2:02=A0am, Chris Ridd <chrisr...@mac.com> wrote:
> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>
> > On Jul 17, 12:22=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> >> On 07/17/10 07:17 PM, hadi motamedi wrote:
> >>> let me know what is wrong here?
>
> >> Only if you stop snipping attributions!
>
> >> --
> >> Ian Collins
>
> > Sorry. What do you mean by 'stop snipping attributions' ?
>
> You tend to leave out the lines I've quoted above (eg On 2010-07-17...)
> which let people see who wrote what.
> --
> Chris

ok, I understand. Can you please comment me back on my problem with
ufsrestore ?

On 07/17/10 09:08 PM, hadi motamedi wrote:
> On Jul 17, 2:02 am, Chris Ridd<chrisr...@mac.com>  wrote:
>> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>>
>>> On Jul 17, 12:22 am, Ian Collins<ian-n...@hotmail.com>  wrote:
>>>> On 07/17/10 07:17 PM, hadi motamedi wrote:
>>>>> let me know what is wrong here?
>>
>>>> Only if you stop snipping attributions!
>>
>>> Sorry. What do you mean by 'stop snipping attributions' ?
>>
>> You tend to leave out the lines I've quoted above (eg On 2010-07-17...)
>> which let people see who wrote what.
>
> ok, I understand. Can you please comment me back on my problem with
> ufsrestore ?

Where's the file you want to restore from?

-- 
Ian Collins

On Jul 17, 3:20=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> On 07/17/10 09:08 PM, hadi motamedi wrote:
>
>
>
> > On Jul 17, 2:02 am, Chris Ridd<chrisr...@mac.com> =A0wrote:
> >> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>
> >>> On Jul 17, 12:22 am, Ian Collins<ian-n...@hotmail.com> =A0wrote:
> >>>> On 07/17/10 07:17 PM, hadi motamedi wrote:
> >>>>> let me know what is wrong here?
>
> >>>> Only if you stop snipping attributions!
>
> >>> Sorry. What do you mean by 'stop snipping attributions' ?
>
> >> You tend to leave out the lines I've quoted above (eg On 2010-07-17...=
)
> >> which let people see who wrote what.
>
> > ok, I understand. Can you please comment me back on my problem with
> > ufsrestore ?
>
> Where's the file you want to restore from?
>
> --
> Ian Collins

I made it on my solaris8 machine as :
#ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
I imported it to my qemu as its virtual disk three with '-hdc
opt8.disk' option so its format command shows disk three on /c0t2d0 .
I tried to restore it on my qemu /opt as :
#ufsrestore -xvf /dev/dsk/c0t2d0s3
But it returns 'volume is not in dump format'

On 2010-07-17 12:04:12 +0100, hadi motamedi said:

> On Jul 17, 3:20�am, Ian Collins <ian-n...@hotmail.com> wrote:
>> On 07/17/10 09:08 PM, hadi motamedi wrote:
>> 
>> 
>> 
>>> On Jul 17, 2:02 am, Chris Ridd<chrisr...@mac.com> �wrote:
>>>> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>> 
>>>>> On Jul 17, 12:22 am, Ian Collins<ian-n...@hotmail.com> �wrote:
>>>>>> On 07/17/10 07:17 PM, hadi motamedi wrote:
>>>>>>> let me know what is wrong here?
>> 
>>>>>> Only if you stop snipping attributions!
>> 
>>>>> Sorry. What do you mean by 'stop snipping attributions' ?
>> 
>>>> You tend to leave out the lines I've quoted above (eg On 2010-07-17...)
>>>> which let people see who wrote what.
>> 
>>> ok, I understand. Can you please comment me back on my problem with
>>> ufsrestore ?
>> 
>> Where's the file you want to restore from?
>> 
>> --
>> Ian Collins
> 
> I made it on my solaris8 machine as :
> #ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
> I imported it to my qemu as its virtual disk three with '-hdc
> opt8.disk' option so its format command shows disk three on /c0t2d0 .
> I tried to restore it on my qemu /opt as :
> #ufsrestore -xvf /dev/dsk/c0t2d0s3
> But it returns 'volume is not in dump format'

ufsdump produces a file (/usr01/opt8.disk in your case). Why are you 
mounting that file as a *disk* in the VM? It isn't a disk.

You might get lucky if you mounted it like you have and then ran 
ufsrestore on the *raw* device, ie ufsrestore -xfv /dev/rdsk/c0t2d0s3. 
But I'm somewhat doubtful that would work.
-- 
Chris

Chris Ridd <chrisridd@mac.com> wrote:
> You might get lucky if you mounted it like you have and then ran 
> ufsrestore on the *raw* device, ie ufsrestore -xfv /dev/rdsk/c0t2d0s3. 
> But I'm somewhat doubtful that would work.

If it were to work at all, he'd need to use the whole-disk slice, s2.  I
don't know why he's using s3.

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On Jul 17, 6:50=A0am, Chris Ridd <chrisr...@mac.com> wrote:
> On 2010-07-17 12:04:12 +0100, hadi motamedi said:
>
>
>
> > On Jul 17, 3:20=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> >> On 07/17/10 09:08 PM, hadi motamedi wrote:
>
> >>> On Jul 17, 2:02 am, Chris Ridd<chrisr...@mac.com> =A0wrote:
> >>>> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>
> >>>>> On Jul 17, 12:22 am, Ian Collins<ian-n...@hotmail.com> =A0wrote:
> >>>>>> On 07/17/10 07:17 PM, hadi motamedi wrote:
> >>>>>>> let me know what is wrong here?
>
> >>>>>> Only if you stop snipping attributions!
>
> >>>>> Sorry. What do you mean by 'stop snipping attributions' ?
>
> >>>> You tend to leave out the lines I've quoted above (eg On 2010-07-17.=
...)
> >>>> which let people see who wrote what.
>
> >>> ok, I understand. Can you please comment me back on my problem with
> >>> ufsrestore ?
>
> >> Where's the file you want to restore from?
>
> >> --
> >> Ian Collins
>
> > I made it on my solaris8 machine as :
> > #ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
> > I imported it to my qemu as its virtual disk three with '-hdc
> > opt8.disk' option so its format command shows disk three on /c0t2d0 .
> > I tried to restore it on my qemu /opt as :
> > #ufsrestore -xvf /dev/dsk/c0t2d0s3
> > But it returns 'volume is not in dump format'
>
> ufsdump produces a file (/usr01/opt8.disk in your case). Why are you
> mounting that file as a *disk* in the VM? It isn't a disk.
>
> You might get lucky if you mounted it like you have and then ran
> ufsrestore on the *raw* device, ie ufsrestore -xfv /dev/rdsk/c0t2d0s3.
> But I'm somewhat doubtful that would work.
> --
> Chris

I import it to my qemu as its third virtual slice with the '-hdc
opt8.disk' option (as its format command shows slice three on /
c0t2d0) . I tried as :
#cd /opt
#ufsrestore -xvf /dev/rdsk/c0t2d0s3
#ufsrestore -xvf /dev/rdsk/c0t2d0s2
But both of them returned as 'volume is not in dump format' .

On Jul 18, 1:22=A0pm, hadi motamedi <motamed...@gmail.com> wrote:
> On Jul 17, 6:50=A0am, Chris Ridd <chrisr...@mac.com> wrote:
>
>
>
>
>
> > On 2010-07-17 12:04:12 +0100, hadi motamedi said:
>
> > > On Jul 17, 3:20=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> > >> On 07/17/10 09:08 PM, hadi motamedi wrote:
>
> > >>> On Jul 17, 2:02 am, Chris Ridd<chrisr...@mac.com> =A0wrote:
> > >>>> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>
> > >>>>> On Jul 17, 12:22 am, Ian Collins<ian-n...@hotmail.com> =A0wrote:
> > >>>>>> On 07/17/10 07:17 PM, hadi motamedi wrote:
> > >>>>>>> let me know what is wrong here?
>
> > >>>>>> Only if you stop snipping attributions!
>
> > >>>>> Sorry. What do you mean by 'stop snipping attributions' ?
>
> > >>>> You tend to leave out the lines I've quoted above (eg On 2010-07-1=
7...)
> > >>>> which let people see who wrote what.
>
> > >>> ok, I understand. Can you please comment me back on my problem with
> > >>> ufsrestore ?
>
> > >> Where's the file you want to restore from?
>
> > >> --
> > >> Ian Collins
>
> > > I made it on my solaris8 machine as :
> > > #ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
> > > I imported it to my qemu as its virtual disk three with '-hdc
> > > opt8.disk' option so its format command shows disk three on /c0t2d0 .
> > > I tried to restore it on my qemu /opt as :
> > > #ufsrestore -xvf /dev/dsk/c0t2d0s3
> > > But it returns 'volume is not in dump format'
>
> > ufsdump produces a file (/usr01/opt8.disk in your case). Why are you
> > mounting that file as a *disk* in the VM? It isn't a disk.
>
> > You might get lucky if you mounted it like you have and then ran
> > ufsrestore on the *raw* device, ie ufsrestore -xfv /dev/rdsk/c0t2d0s3.
> > But I'm somewhat doubtful that would work.
> > --
> > Chris
>
> I import it to my qemu as its third virtual slice with the '-hdc
> opt8.disk' option (as its format command shows slice three on /
> c0t2d0) . I tried as :
> #cd /opt
> #ufsrestore -xvf /dev/rdsk/c0t2d0s3
> #ufsrestore -xvf /dev/rdsk/c0t2d0s2
> But both of them returned as 'volume is not in dump format' .

Because you are not using your dumped file.  Take the file and COPY it
somewhere on
your machine you can access.

cd /opt
ufsrestore -rvf /opt8.disk (or where ever you have it stored.)

Before doing that, CHECK THE MAN PAGE,  I don't have a machine
powered on to check this at the moment.

On Jul 18, 3:41=A0pm, chuckers <chucker...@gmail.com> wrote:
> On Jul 18, 1:22=A0pm, hadi motamedi <motamed...@gmail.com> wrote:
>
>
>
> > On Jul 17, 6:50=A0am, Chris Ridd <chrisr...@mac.com> wrote:
>
> > > On 2010-07-17 12:04:12 +0100, hadi motamedi said:
>
> > > > On Jul 17, 3:20=A0am, Ian Collins <ian-n...@hotmail.com> wrote:
> > > >> On 07/17/10 09:08 PM, hadi motamedi wrote:
>
> > > >>> On Jul 17, 2:02 am, Chris Ridd<chrisr...@mac.com> =A0wrote:
> > > >>>> On 2010-07-17 09:45:29 +0100, hadi motamedi said:
>
> > > >>>>> On Jul 17, 12:22 am, Ian Collins<ian-n...@hotmail.com> =A0wrote=
:
> > > >>>>>> On 07/17/10 07:17 PM, hadi motamedi wrote:
> > > >>>>>>> let me know what is wrong here?
>
> > > >>>>>> Only if you stop snipping attributions!
>
> > > >>>>> Sorry. What do you mean by 'stop snipping attributions' ?
>
> > > >>>> You tend to leave out the lines I've quoted above (eg On 2010-07=
-17...)
> > > >>>> which let people see who wrote what.
>
> > > >>> ok, I understand. Can you please comment me back on my problem wi=
th
> > > >>> ufsrestore ?
>
> > > >> Where's the file you want to restore from?
>
> > > >> --
> > > >> Ian Collins
>
> > > > I made it on my solaris8 machine as :
> > > > #ufsdump 0f /usr01/opt8.disk /dev/dsk/c0t0d0s3
> > > > I imported it to my qemu as its virtual disk three with '-hdc
> > > > opt8.disk' option so its format command shows disk three on /c0t2d0=
 .
> > > > I tried to restore it on my qemu /opt as :
> > > > #ufsrestore -xvf /dev/dsk/c0t2d0s3
> > > > But it returns 'volume is not in dump format'
>
> > > ufsdump produces a file (/usr01/opt8.disk in your case). Why are you
> > > mounting that file as a *disk* in the VM? It isn't a disk.
>
> > > You might get lucky if you mounted it like you have and then ran
> > > ufsrestore on the *raw* device, ie ufsrestore -xfv /dev/rdsk/c0t2d0s3=
..
> > > But I'm somewhat doubtful that would work.
> > > --
> > > Chris
>
> > I import it to my qemu as its third virtual slice with the '-hdc
> > opt8.disk' option (as its format command shows slice three on /
> > c0t2d0) . I tried as :
> > #cd /opt
> > #ufsrestore -xvf /dev/rdsk/c0t2d0s3
> > #ufsrestore -xvf /dev/rdsk/c0t2d0s2
> > But both of them returned as 'volume is not in dump format' .
>
> Because you are not using your dumped file. =A0Take the file and COPY it
> somewhere on
> your machine you can access.
>
> cd /opt
> ufsrestore -rvf /opt8.disk (or where ever you have it stored.)
>
> Before doing that, CHECK THE MAN PAGE, =A0I don't have a machine
> powered on to check this at the moment.

Excuse me, I cannot transfer the 'opt8.disk' to my qemu as a file .
What I have here is to import it as its third virtual slice with the '-
hdc opt8.disk' option . Then the format output shows slice three on /
c0t2d0 that I need to restore on my qemu /opt .

hadi motamedi <motamedi24@gmail.com> wrote:
> Excuse me, I cannot transfer the 'opt8.disk' to my qemu as a file .
> What I have here is to import it as its third virtual slice with the '-
> hdc opt8.disk' option . Then the format output shows slice three on /

It isn't possible to import a file as a slice on a virtual hard disk.
It is either the disk or it isn't.  Which means it is either the WHOLE
disk or it's not.  It isn't possible to import it as the "third virtual
slice".  I don't know how you're concluding this, but you're wrong.

s2 is normally the whole disk, assuming you haven't edited the slices on
the virtual disk... which will damage the underlying dump file.

I'm reviewing the qemu documentation... when you say "third virtual slice",
do you really mean "third virtual disk", which is what -hdc means?  Slices
are NOT disks.

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On Jul 19, 5:27=A0am, hume.spamfil...@bofh.ca wrote:
> hadi motamedi <motamed...@gmail.com> wrote:
> > Excuse me, I cannot transfer the 'opt8.disk' to my qemu as a file .
> > What I have here is to import it as its third virtual slice with the '-
> > hdc opt8.disk' option . Then the format output shows slice three on /
>
> It isn't possible to import a file as a slice on a virtual hard disk.
> It is either the disk or it isn't. =A0Which means it is either the WHOLE
> disk or it's not. =A0It isn't possible to import it as the "third virtual
> slice". =A0I don't know how you're concluding this, but you're wrong.
>
> s2 is normally the whole disk, assuming you haven't edited the slices on
> the virtual disk... which will damage the underlying dump file.
>
> I'm reviewing the qemu documentation... when you say "third virtual slice=
",
> do you really mean "third virtual disk", which is what -hdc means? =A0Sli=
ces
> are NOT disks.
>
> --
> Brandon Hume =A0 =A0- hume -> BOFH.Ca,http://WWW.BOFH.Ca/

Thank you very much for your reply. But the /c0t2d0s2 cannot be
mounted as well .

hadi motamedi <motamedi24@gmail.com> wrote:
> Thank you very much for your reply. But the /c0t2d0s2 cannot be
> mounted as well .

Of course it can't... you said the underlying file was a DUMP file, not
a disk image.  chuckers asked you earlier... why are you trying to treat
a dump file as a disk?  It isn't a disk.  It's a dump file.  It does NOT
contain a filesystem you can mount... it contains a dump file meant to
be written to tape.  It can't be mounted, ever, no matter what you do.

What does "ufsrestore -tf /dev/rdsk/c0t2d0s2" say?

And let me say this again: Everything you've done has been the absolute
worst way to accomplish anything.  Someone with your skill level should have
purchased (or gotten your employer to purchase) a small, cheap SPARC unit.

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On Jul 19, 5:53=A0am, hume.spamfil...@bofh.ca wrote:
> hadi motamedi <motamed...@gmail.com> wrote:
> > Thank you very much for your reply. But the /c0t2d0s2 cannot be
> > mounted as well .
>
> Of course it can't... you said the underlying file was a DUMP file, not
> a disk image. =A0chuckers asked you earlier... why are you trying to trea=
t
> a dump file as a disk? =A0It isn't a disk. =A0It's a dump file. =A0It doe=
s NOT
> contain a filesystem you can mount... it contains a dump file meant to
> be written to tape. =A0It can't be mounted, ever, no matter what you do.
>
> What does "ufsrestore -tf /dev/rdsk/c0t2d0s2" say?
>
> And let me say this again: Everything you've done has been the absolute
> worst way to accomplish anything. =A0Someone with your skill level should=
 have
> purchased (or gotten your employer to purchase) a small, cheap SPARC unit=
..
>
> --
> Brandon Hume =A0 =A0- hume -> BOFH.Ca,http://WWW.BOFH.Ca/

Thank you very much for your help. As I have understood from your
message , my procedure is wrong . Can you please let me know if you
have experienced importing one of your folders from your solaris
machine to qemu-system-sparc ?

hadi motamedi <motamedi24@gmail.com> wrote:
> message , my procedure is wrong . Can you please let me know if you
> have experienced importing one of your folders from your solaris
> machine to qemu-system-sparc ?

I have never used qemu for anything.  I've always used a *real* SPARC
machine.

You need to talk to the people on the qemu forums.

-- 
Brandon Hume    - hume -> BOFH.Ca, http://WWW.BOFH.Ca/

On Jul 20, 5:28=A0am, hume.spamfil...@bofh.ca wrote:
> hadi motamedi <motamed...@gmail.com> wrote:
> > message , my procedure is wrong . Can you please let me know if you
> > have experienced importing one of your folders from your solaris
> > machine to qemu-system-sparc ?
>
> I have never used qemu for anything. =A0I've always used a *real* SPARC
> machine.
>
> You need to talk to the people on the qemu forums.
>
> --
> Brandon Hume =A0 =A0- hume -> BOFH.Ca,http://WWW.BOFH.Ca/

Thank you very much for your help.