HELP! Return value 12 - system will not boot.
Rob
to a network setting. The system locked up, and would not respond. I
have actually never seen Unixware do this before. I let it sit for a
while, and no luck, so I powered it down and back up. Now I get the
following error after the splash screen:
UX:/sbin/ckroot: WARNING: Return value 12
UX:/sbin/ckroot: INFO: system will reboot
Press any key to restart
And that's it, no other error messages at all. I can't do anything,
and I can't find anything about this or any "return values" on sco's
website.
Any help would be greatly appreciated.
Rob
Matt Schalit
Can you boot from emergency disks and fsck the root
filesystem?
What version of UnixWare?
What's your hardware and HBA configuration?
So you do get a splash screen?
Can you hit the spacebar and go to the
boot: promt and enter single user mode?
Good Luck,
Matt
Rob
Matt,
This is a new system, and I have not yet created the emergency repair
disks.
It is Unixware 7.10.
System is a Dell Poweredge 2500 - Dual ghz - 1gig RAM - (3) 36gig SCSI
hard drives on Raid 5 with a Perc3 raid controller.
Yes I get a splash screen, and I can get to the boot prompt, but what
is the command to get into single user mode from there?
Thanks,
Rob
Tony Lawrence
> Yes I get a splash screen, and I can get to the boot prompt, but what
> is the command to get into single user mode from there?
If I remember correctly, you can type "?" to get a list of valid
commands.
In fact, I'm pretty sure it says exactly that that :-)
but anyway: for single user mode, type:
INITSTATE=S
boot
--
Tony Lawrence
SCO/Linux Support Tips, How-To's, Tests and more: http://pcunix.com
Rob
Thanks,
Rob
Matt Schalit
Hi Rob,
For us to get somewhere, you're going to want
to take the time to explicitly type the facts
for us to understand. Your reply is brief.
What we are going to need is what you did and
what you saw.
Here's what I'll guess. When you got to the
boot prompt, you entered what Tony told you,
and something failed.
Please tell us exactly what you saw on the
screen after you typed
initstate=s
boot
You can abbreviate any licensing, copyright,
or registration text.
Either your kernel loaded or it didn't.
We'll find that out from what you saw.
What sort of tape backup device do you have?
Do you have a full backup?
Good Luck,
Matthew
Rob
Matthew,
After typing
initstate=s
boot
the system returned to the splash screen.
I get the copyright notice.
A system memory readout.
"The system is coming up, please wait".
A readout of the drives attached to my system.
A "NOTICE that my system RAM is larger than the dump device, so system
dump will be selective" - I have always gotten this, even on my other
system that has been up and running for 2 years.
a "INFO: The system is being examined, this may take a few minutes."
Then I get the error I have listed before:
UX:/sbin/ckroot: WARNING: Return value 12
UX:/sbin/ckroot: INFO: system will reboot
Press any key to restart
All of the listed above happens within a matter of seconds.
Thanks for your help, anyone has has an idea, please let me know.
Thanks,
Rob
Roberto Zini
According to the man page, ckroot reads the mount options for
the root filesystem from the /etc/vfstab file; if the error
returned is coherent with the sys/errno.h file, 12 should
mean ENOMEM (not enough core ?).
Perhaps the /etc/vfstab file's got screwed up; are you in the
position to boot from the emergency floppies set and check
that file (or even better, restore it from backup) ?
Best,
Roberto
--
---------------------------------------------------------------------
Roberto Zini email : r.z...@strhold.it
Technical Support Manager -- Strhold Evolution Division R.E. (ITALY)
---------------------------------------------------------------------
"Has anybody around here seen an aircraft carrier?"
(Pete "Maverick" Mitchell - Top Gun)
Rob
Roberto,
Thanks for the reply. This is a new server, and I have not yet made
the emergency repaid disks, nor any backups. I still have another
server I am running live on, but it has different hardware and
filesystem setups. Is there any other way to repair these files?
Thanks again.
Rob
Matt Schalit
>Roberto Zini <r.z...@strhold.it> wrote in message news:<3BCAF173...@strhold.it>...
>> Rob wrote:
>> >
>> > msch...@pacbell.net (Matt Schalit) wrote in message news:<3bc74eb7...@news.sf.sbcglobal.net>...
>> > Matthew,
>> >
>> > After typing
>> >
Getting to the boot: prompt means that
you Master Boot Block, MBB, sector 0
on you hd has been successfully read and
has been told the active partition, which
has been found, whose secondary bootstrap
program has found the boot slice, /stand,
and loaded the raw data off there, which are
many .blm files, like stage3.blm. Also
loaded is /stand/bootmsgs and /stand/boot,
and you see the boot: prompt.
>> > initstate=s
>> > boot
Ok.
>> > the system returned to the splash screen.
>> >
>> > I get the copyright notice.
>> >
>> > A system memory readout.
Apparently all the memory was found, too,
or you'd mention that, I figure...
>> > "The system is coming up, please wait".
Aha. Your /stand/unix kernel has been successfully
loaded. You have good kernel.
>> > A readout of the drives attached to my system.
a. Sysproc (as PID 0) loads, executes, and starts init.
/root, /var are mounted for reading.
b. Init (as PID 1) executes, and reads inittab to set the initial
run level and mount filesystems.
The first three lines of inittab are:
fr::sysinit:/sbin/fixroot >/dev/sysmsg 2>&1
swp1::sysinit:/sbin/swap -a /dev/swap >/dev/sysmsg 2>&1
cr::sysinit:/sbin/ckroot >/dev/sysmsg 2>&1
I don't know what fixroot does.
The swap line may be related to this output.
>> > A "NOTICE that my system RAM is larger than the dump device, so system
>> > dump will be selective" - I have always gotten this, even on my other
>> > system that has been up and running for 2 years.
That message, though, definitely goes away once your system
is patched with set71101i and a few other ptfs.
Finally, ckroot_vxfs(1M) gets run by that /sbin/ckroot
command, and reads /etc/vfstab to determine what to do about
the filesystems it's supposed to mount rw.
It's supposed to ckeck and clean ones that are marked
dirty.
>> > a "INFO: The system is being examined, this may take a few minutes."
>> > Then I get the error I have listed before:
>> >
>> > UX:/sbin/ckroot: WARNING: Return value 12
>> > UX:/sbin/ckroot: INFO: system will reboot
>> > Press any key to restart
Comments below....
>> > All of the listed above happens within a matter of seconds.
>> >
>> > Thanks for your help, anyone has has an idea, please let me know.
>> >
>> > Thanks,
>> >
>> > Rob
>>
>> According to the man page, ckroot reads the mount options for
>> the root filesystem from the /etc/vfstab file; if the error
>> returned is coherent with the sys/errno.h file, 12 should
>> mean ENOMEM (not enough core ?).
Let's be clear on this for grins, man 2 intro:
'12 ENOMEM Insufficient or invalid memory
During execution of an exec, brk, or sbrk routine,
a program asks for more space than the system is able
to supply. This is not a temporary condition; the
maximum size is a system parameter. The error may also
occur if the arrangement of text, data, and stack
segments requires too many segmentation registers, or
if there is not enough swap space during the fork routine.
...'
>> Perhaps the /etc/vfstab file's got screwed up;
That's the inidication, as it's possible that the
swp1::sysinit:/sbin/swap -a /dev/swap >/dev/sysmsg 2>&1
command did not succesfully complete. If so, then there
would be no swap space.
>> are you in the
>> position to boot from the emergency floppies set and check
>> that file (or even better, restore it from backup) ?
>>
>> Best,
>> Roberto
>>
>Roberto,
>
>Thanks for the reply. This is a new server, and I have not yet made
>the emergency repaid disks, nor any backups. I still have another
>server I am running live on, but it has different hardware and
>filesystem setups. Is there any other way to repair these files?
>
>Thanks again.
>
>Rob
I'm don't know how to repair a damaged vfstab
file off the top of my head. Anybody else know?
To get access to the system, you can slap the
hard disk in the other Uw7 machine (as long as
the scsi id's don't clash), then fsck the slices.
Perhaps you can't do that. Another approach
would be to create a "Maigc Floppy." You remeber
that one don't you? It's the floppy disk that the
ISL instructions said you didn't need :)
Use a Uw7 or Windoze mochine to create the diskette.
Mount the cdrom to /mnt, then cd to /mnt/info/images
where you'll find the Magic image. Then use the
two ISL diskettes to boot the broken uw7 machine like
you were going to do an ISL. When the boot process
gets to the point where it asks you to load vendor
supplied HBA diskettes, load the Magic Floppy. Then
swith to vt0 and you'll have a prompt where you can
chroot and use vi, fsck, and a few others.
If you want to examine the contents of the Magic
Floppy before you try to use it, then you can mount
it on the other Uw7 machine. It's an s5 filesystem:
mount -F s5 /dev/fd0 /mnt
You'll run into a point where doing a lot of research
into this and fixing this (over the last few days
included) will cost more of your valuable time than
a fresh install.
Good Luck,
Matt
Rob
> file off the top of my head. Anybody else know?
Anyone with any ideas?
Thanks,
Rob
Roberto Zini
Hi !
I know this is NOT what you'd have expected to hear but I see
that this thread is running since last Thursday (ie, a whole
week); wouldn't have been better for you to reinstall the whole
box from scratch ?
Depending on the HW being used, a complete UW7 installation should
take approx 2 hours to complete and since you stated that it was
a new server, you could have a completely working machine by now.
A apologize in advance if this seems a lot "crude", but given
the lack of ideas I've got at present, this is the quickest solution
I can think of.
Serge Bromow
news:a021963f.01101...@posting.google.com...
> Anyone with any ideas?
>
> Thanks,
>
> Rob
Since you have no backups and no boot disk
and apparently a lot of time;
JUST REINSTALL THE DAM OS.
This time add a whole bunch of swap.
--
Posted from port-31.ottawa4.achilles.net [209.151.2.130]
via Mailgate.ORG Server - http://www.Mailgate.ORG
Rob
> "Rob" <rab...@thecommercegroup.com> wrote in message
> news:a021963f.01101...@posting.google.com...
>
> > Anyone with any ideas?
> >
> > Thanks,
> >
> > Rob
>
>
> Since you have no backups and no boot disk
> and apparently a lot of time;
>
> JUST REINSTALL THE DAM OS.
>
> This time add a whole bunch of swap.
Well guys, here's the deal.
Point #1: This is the second time I have gotten this error since this
new server has been in my building. I have loaded the OS TWICE, and I
would really like to know what is causing the error so I don't get it
again when I go live with it!!! I know how long it takes to load UW7,
but I want to know the problem, not the easy solution. If this
happens again when the server is in production, I can't just keep
re-loading everything.
Point #2: People are suggesting it is running out of swap file space.
How can this be? I have our older UW7 server sitting right next to
this new one, and I have NEVER had anything like this in the past 2
years. This new server is setup EXACTLY the same way as the old one,
except for the fact that it is newer, faster hardware, and that I
doubled all the filesystem sizes including the one where the swap file
lives. So if I have two systems that have all the same data on them,
running the same SINGLE application, then how come the new one is
running out of swap file space???
I appreciate everyone's help, but I need an answer to the cause, not a
"relaod it" solution. If I can't find it here, then I'll have to go
pay mega $$ for some SCO support person to tell me to re-load it.
Thanks,
Rob
Ronald J Marchand
> "Serge Bromow" <se...@omensys.com> wrote in message
news:<700e9753157e6847f29...@mygate.mailgate.org>...
> > "Rob" <rab...@thecommercegroup.com> wrote in message
> > news:a021963f.01101...@posting.google.com...
> >
> > > Anyone with any ideas?
Have you tried replacing hardware items?
Matt Schalit
>"Serge Bromow" <se...@omensys.com> wrote in message news:<700e9753157e6847f29...@mygate.mailgate.org>...
>> "Rob" <rab...@thecommercegroup.com> wrote in message
>> news:a021963f.01101...@posting.google.com...
>>
>> > Anyone with any ideas?
>> >
>> > Thanks,
>> >
>> > Rob
>>
>>
>> Since you have no backups and no boot disk
>> and apparently a lot of time;
>>
>> JUST REINSTALL THE DAM OS.
>>
>> This time add a whole bunch of swap.
>
>Well guys, here's the deal.
>
>Point #1: This is the second time I have gotten this error since this
>new server has been in my building. I have loaded the OS TWICE, and I
>would really like to know what is causing the error so I don't get it
>again when I go live with it!!! I know how long it takes to load UW7,
>but I want to know the problem, not the easy solution. If this
>happens again when the server is in production, I can't just keep
>re-loading everything.
Understandable. We're hoping you can just restore
your vfstab or more from backup tapes, but we figure
you'd have done that if you had any.
I suppose it's time to hear about what your
hardware is, and what drivers you're using from
ftp://ftp.caldera.com/pub/unixware7/drivers
>Point #2: People are suggesting it is running out of swap file space.
> How can this be?
The vfstab contains the settings for the swap file device
to be created at boot time. Your vfstab is corrupt. You
therefore can't succesfully initialize your swap.
I have our older UW7 server sitting right next to
>this new one, and I have NEVER had anything like this in the past 2
>years. This new server is setup EXACTLY the same way as the old one,
>except for the fact that it is newer, faster hardware, and that I
>doubled all the filesystem sizes including the one where the swap file
>lives. So if I have two systems that have all the same data on them,
>running the same SINGLE application, then how come the new one is
>running out of swap file space???
>
>I appreciate everyone's help, but I need an answer to the cause, not a
>"relaod it" solution. If I can't find it here, then I'll have to go
>pay mega $$ for some SCO support person to tell me to re-load it.
>
>Thanks,
>
>Rob
Most likely a patch/driver issue. As you mention,
Uw7 tends to run great when you've stablized it to
your hardware.
Matt