[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: debian hangs during boot (at fsck stage)



> 
> Shaul Karl wrote:
> > 
> > >
> > > Sorry, should have done some reading before posting the above, my bad...
> > >
> > > I've read the LILO howto and that says that pressing tab should do the trick,
> > > or holding down alt or shift when the LILO bit comes up. But neither of these
> > > action seem to make any difference - the machine continues booting regardless.
> > >
> > > thanks
> > > alex
> > >
> > 
> > I believe not being able to enter text at the lilo prompt has something
> > to do with your lilo configuration but I do not know more then that. I
> > do not know about the MBR thing either.
> > 
> > An alternative strategy might be to boot from a rescue disk and try to
> > continue from there.
> 
> I've tried this and get "kernel panic: could not mount root file system." 
> A screenshot is at 
> http://www.tardis.ed.ac.uk/~lard/kernel.panic.jpg
> 
> I notice it mentions SCSI things again, but I've not installed any SCSI
> devices.
> 


It too notices you do not have SCSI devices.
I Believe it get stacked with the first partition on the first IDE 
device, that is hda1. As far as I know device 03:01 is hda1. What have 
you got there?
Can you tell how come the installed system did not have any problems 
with this device? Perhaps you need another rescue disk? Maybe one with 
a kernel patched with UDMA? It does identify your HDs correctly, 
doesn't it?
Perhaps this is a matter of telling the kernel it should use another 
root device? Is that what the rdev command is used for? If I am not 
mistaken you can pass the rescue floppy kernel a parameter with this 
information.


> What should I do in this situation? I'm more concerned with not losing data on
> the disk than resucing my debian install, although doing both would be
> preferable.
> 


Another method might be to use a boot+root floppies as if you are going 
to install, quit the installation as soon as you have a usable 2nd VT 
and continue from there. More specifically, use the appropriate 
floppies as if you are going to install. Let the installation process 
initialize your kbd and perhaps one or 2 other things it does 
initially, and then switch to the 2nd VT. Hopefully now you can mount 
your partitions, at least those that are mountable, and use the 
debug/repair tools that are there.
I believe that a similar process can be followed with a bootable CD.


> > As for the file system debug/repair tools I referred to, I meant e2fsck
> > and recursively all the other programs that are mentioned in the `See
> > Also' part of its manpage. If the bad partition is not ext2 then there
> > should be similar programs for the relevant file system.
> > When dealing with those programs you might want to compare the output
> > of a well behaved partition to that of the bad one.
> > Of course, these programs should be handled with extra care so that you
> > will not cause damage to a healthy partition.
> 
> Ok, thanks for that info.  
> 
> alex

-- 
When responding, please quote my entire message.

    Shaul Karl <shaulka@bezeqint.net>




Reply to: