fsck on sco 5.0.5 open server. where is it?
romppast
-----------------------------------------------------------------------
----------
Im new to SCO server family and recently just encountered problem with
my SCO 5.0.5 Openserver as it hangs-up while booting with the message of
**Root File System Modified** followed by **Remounted Root File System**
then stops from there. This happened after the server was turned-off
accidentally by my baby :) while i was doing a vi on resolv.conf.
The bad thing is, i bought the server second hand without installer
disks, boot/root floppies and manual to look up to so im just relying on
the internet in finding cure to this problem.
I check on google and found infos on how to boot the system using
oss604a.boot and from there mount /dev/hd0root /mnt. However, in most of
the advices on threads, fsck is required but cant call it from the /etc
and not even in the mounted /mnt/bin as it is not there. Im wondering
why didnt they put this fsck on this boot disk...
Im starting to lose patience and contemplating on replacing the os with
freedbsd or linux as i found them very easy to handle with tons of
infos, fixes and organized manuals available over the internet. Im at
lost and dont sure if this is the right move to take.
Until i found myself this morning looking up the sky along with my baby
as if she too is looking for an answer to my problem, with a state of
mind you may refer to as "buffer overflow" when it came to my mind that
there is this forum that may be of help to me.
Please help me out with this...thanks and more power to all of you guys
--
Posted via http://dbforums.com
Bela Lubkin
> Im new to SCO server family and recently just encountered problem with
> my SCO 5.0.5 Openserver as it hangs-up while booting with the message of
> **Root File System Modified** followed by **Remounted Root File System**
> then stops from there. This happened after the server was turned-off
> accidentally by my baby :) while i was doing a vi on resolv.conf.
>
> The bad thing is, i bought the server second hand without installer
> disks, boot/root floppies and manual to look up to so im just relying on
> the internet in finding cure to this problem.
>
> I check on google and found infos on how to boot the system using
> oss604a.boot and from there mount /dev/hd0root /mnt. However, in most of
> the advices on threads, fsck is required but cant call it from the /etc
> and not even in the mounted /mnt/bin as it is not there. Im wondering
> why didnt they put this fsck on this boot disk...
If you're getting those messages, the system is running fsck for you, so
it's certainly present. I think you would only get those messages about
the root filesystem after a question like "Root filesystem is dirty, OK
to check it?". One thing you can do is: say no. Then go to single-user
mode and run:
fsck -o full /dev/root
and see how it goes.
Using the boot floppy, you would need to do this:
mount /dev/hd0root /mnt
chroot /mnt /bin/sh
fsck -o full /dev/root
But I think this won't work because of the way root is mounted. fsck is
only willing to check a mounted filesystem if it's the root filesystem
(according to the booted kernel). When booted from a floppy, the floppy
is the root. You _might_ be able to:
mount -r /dev/hd0root /mnt
chroot /mnt /bin/sh
fsck -o full /dev/root
Doing it from single-user mode is probably much easier.
If you're still stuck, post the full messages you're getting during the
failed bootup.
>Bela<
romppast
hi Bela,
Thanks to your prompt reply to my post this morning.
I've tried your suggestion as follows:
OPTION 1:
Boot and answered no to "root filesstem is dirty, OK to check it?".
There was no shell prompt to write the next command; fsck -o full
/dev/root; on so instead i just went ahead and type the command on the
blank line.
I dont know if it was just coincidence or i did it right, the hard
disk's LED flickered continuously but no activities appearing on
the monitor.
20minutes after, the LED activity seemed to be not right as it just
flickered on and on and in constant interval as if a microsoft windows
scans a hd bad sector.
Is this normal? how long normally an fsck lasts witha 40GB hd.
OPTION 3:
Using the floppy, i was able to do the mknod /dev/hd0root b 1 42, mount
-r /dev/hd0root /mnt but failed with the chroot /mnt /bin/sh with an
error of bad address.
I did, cd /mnt/bin, echo * and showed STTY]...chmode, chroot cp...
I tried the rest of the commands such as the cp, fdisk, vi, cat by
calling them as is or by entering /mnt/bin/chroot but also resulted with
the same error message.
I deeply appreciate your help.
Best regards,
Rommel P. Pastrana
Originally posted by Bela Lubkin
--
Posted via http://dbforums.com
Bela Lubkin
> I've tried your suggestion as follows:
>
> OPTION 1:
>
> Boot and answered no to "root filesstem is dirty, OK to check it?".
> There was no shell prompt to write the next command; fsck -o full
> /dev/root; on so instead i just went ahead and type the command on the
> blank line.
Normally, after answering "no" to that question, you would get a prompt
asking you to type the root password to go into single-user mode, or ^D
to go multi-user. You would have entered the root password, gotten a
shell prompt, then run the fsck.
I suppose it is _possible_ that you did actually get to a shell prompt
that wasn't displaying (for some strange reason). One thing you can do
to see if you're at a shell prompt is type "echo ^G" (control-G). It
may beep when you hit ^G; it may beep when you hit return and it acts on
the command. If it beeps when you hit return, you are definitely at
some sort of shell prompt. If it doesn't beep, the test is inconclusive
(the bell character might have disappeared for the same reason the rest
of the shell output is invisible).
> I dont know if it was just coincidence or i did it right, the hard
> disk's LED flickered continuously but no activities appearing on
> the monitor.
>
> 20minutes after, the LED activity seemed to be not right as it just
> flickered on and on and in constant interval as if a microsoft windows
> scans a hd bad sector.
>
> Is this normal? how long normally an fsck lasts witha 40GB hd.
That actually sounds like a reasonable description of disk light
activity during an `fsck -o full`. fsck has various phases it goes
through, which look different. Of course the whole thing should end
eventually. Total time depends on a bunch of factors -- speed of the
hard disk (mostly seek speed), number of files, amount of space that's
in use, etc. A full 40GB drive would probably take a long time, 20
minutes does not sound out of range.
You didn't say how this ended -- did you hit RESET after deciding it
must be hung? Maybe try it again, but let it run longer -- go to
lunch...
> OPTION 3:
>
> Using the floppy, i was able to do the mknod /dev/hd0root b 1 42, mount
> -r /dev/hd0root /mnt but failed with the chroot /mnt /bin/sh with an
> error of bad address.
Please give exact output when you report an error message. In fact,
it's best if you show the command and result as accurately as possible,
something like:
The chroot failed like this:
# chroot /mnt /bin/sh
Bad address
> I did, cd /mnt/bin, echo * and showed STTY]...chmode, chroot cp...
>
> I tried the rest of the commands such as the cp, fdisk, vi, cat by
> calling them as is or by entering /mnt/bin/chroot but also resulted with
> the same error message.
I'm not sure what to make of that message, don't have enough context.
In your first message you wrote:
> > > The bad thing is, i bought the server second hand without installer
> > > disks, boot/root floppies and manual to look up to so im just
> > > relying on the internet in finding cure to this problem.
It begins to seem like the system you bought was sick in the first
place, somehow improperly installed. I can't think of what would cause
the symptoms you describe. It is probably repairable, but repair will
be very difficult when every step has to be discussed over several days
on USENET.
>Bela<
romppast
Hi Bela,
yes, the actual message was:
#chroot /mnt /bin/sh
Bad address
#/mnt/bin/chroot /mnt /bin/sh
Bad address
another command attempt was:
#fsck -o full /dev/root
Bad address
Alright, I will try fsck longer.
If in case the above fails, will a boot/root floppy taken from an sco
5.0.5 server help, just to get to an actual shell and run fsck? If so,
where can i download an emergency root file system floppy as i already
have the boot?
Are there SCO 5.0.5 Demo Installer that I can download so i can try
reinstalling it or another option of maybe doing a fresh install to a
new hard disk and then making the troubled one slave in order for me to
recover still some applications and data from it.
Again, thanks so much for your help.
Regards.
Rommel P. Pastrana
Bill Vermillion
But if he can't see it he won't know when it ends.
Maybe he should try entering the fsck -ofull as
fsck -ofull; /etc/shutdown -g0 -y -i<n> with n
being 0 for shutdown, 1 for single user, or 6 for reboot.
Then do something else for an hour or two.
Does that sound like a reasonable approach for an unreadable
screen? Perhaps checking the video options in the BIOS might be
good too. Not all BIOS support that but I made one workable - but
a bit slow until I got done but setting the video to VGA.
Bill
--
Bill Vermillion - bv @ wjv . com
Nachman Yaakov Ziskind
Maybe he should try
# fsck -ofull -y
On the theory that the display is hosed, and fsck might be asking some
questions ...
NYZ
--
_________________________________________
Nachman Yaakov Ziskind, EA, LLM aw...@egps.com
Attorney and Counselor-at-Law http://yankel.com
Economic Group Pension Services http://egps.com
Actuaries and Employee Benefit Consultants
romppast
Hi to all,
This is what i did this afternoon in the same order:
1. fsck -ofull; /etc/shutdown -g0 -y -i
_
result: just a cursor appeared following the command, ^G produces beep,
pressing enter causes the text to go up by 1 line, ^D no beep, HD led is
not flickering, seemed like no activity.
2. fsck -ofull -y
result: same as above; 1.
3. Via Floppy
#mknod /dev/hd0root b 1 42
#mount -r /dev/hd0root /mnt
#chroot /mnt /bin/sh
chroot:not found
#fsck -ofull /dev/hd0root
fsck:not found
4. fsck -ofull; /etc/shutdown -g0 -y -i
_
result: tomorrow morning; 8 hours after.
If the 4th move fails, will an emergency boot/root floppy; with an fsck;
helps? Are there users in here; or sites that you know; who are willing
to share a root file system taken from an SCO5.0.5. I will be willing to
pay for it as long as it wont be that much, you know :). But if you guys
believe that that wont work either then that will be it, case close,
good bye, sayonara, asta lavista SCO 5.0.5.
Again, thank you so much Bela, Bill and Nachman for sharing your
thoughts and i wish i could return the favor someday.
Best regards,
Rommel P. Pastrana
--
Posted via http://dbforums.com
Nachman Yaakov Ziskind
Does it violate my SCO/Microlite license agreements to share my RecoverEdge
media? If not, I'd be happy to email you images (you put them on floppies with
rawrite.exe, a program you can d/l). From 5.0.6, and, no, I have no idea if
they can help (different hardware, etc.) But you *can* break out to a shell
prompt with them, AFAIK, and the environment there is more friendly ...
romppast
Correction, i should have written the follwing with test 3 (via floppy);
Boot with floppy then press F8 when install media is being asked.
Then,
<installation>mknod /dev/hd0root b 1 42
<installation>mount -r /dev/hd0root /mnt
<installation>chroot /mnt /bin/sh
<installation>chroot:not found
<installation>fsck -o full /dev/hd0root
<installation>fsck:not found
On test 4. No result after a 10hours. Display is still:
fsck -ofull; /etc/shutdown -g0 -y -i
_
Thanks Nachman, how nice of you to offer your recovery image. Yes, i
understand about this license issues, things that are so true with
SCO. I wish SCO will be friendly on fixes like this as i believe these
are just some of the basic things that they have to make available on
their site.
I dont want to compare but I cant help to think of it.
Thanks and more power to dbforum.
Rommel