[Fedora 12] "IOError: [Erno 28] No space left on device"

[Fernando Cassia]

I have one old machine running Fedora 12, kernel 2.6.27.
Suddenly one day after doing system updates (and it was downloading
some torrents in the background) the system would no longer boot
properly, and the system says the disk is full (probably is).

The issue is that I get the normal Fedora boot screen , then the
hourglass, then the log-in screen,then I get

Oops, sealert hit an error!

Traceback (most recent call last):

File "/usr/bin/sealert*, line 949, in <module>
  print >> sys.stderr, "could not attach to desktop process"
IOError: [Erno 28] No space left on device

This dialog shows an "ok" button. I click there and ... nothing...
black screen and spinning hourglass...

I´ve tried Ctrl-Alt-F1 but nothing happens, Ctrl-Alt-F2 shows

IRQ handler tye mismatch for IRQ 0
current handler: timer
IRQ handler tye mismatch for IRQ 0
current handler: timer
IRQ handler tye mismatch for IRQ 0
current handler: timer
IRQ handler tye mismatch for IRQ 0
current handler: timer
hub 1-0:1.0: unable to enumerate USB device on port 1

Any ideas?? I´m lost on this one...

TIA
FC
PS: I don´t have a functioning CD/DVD reader on this system. I used at
the time of installation but not anymore...

[Kevin Cummings]

On 02/27/2011 05:01 PM, Fernando Cassia wrote:
>  I have one old machine running Fedora 12, kernel 2.6.27.
>  Suddenly one day after doing system updates (and it was downloading
>  some torrents in the background) the system would no longer boot
>  properly, and the system says the disk is full (probably is).
>  
>  The issue is that I get the normal Fedora boot screen , then the
>  hourglass, then the log-in screen,then I get
>  
>  Oops, sealert hit an error!
>  
>  Traceback (most recent call last):
>  
>  File "/usr/bin/sealert*, line 949, in <module>
>    print >> sys.stderr, "could not attach to desktop process"
>  IOError: [Erno 28] No space left on device
>  
>  This dialog shows an "ok" button. I click there and ... nothing...
>  black screen and spinning hourglass...
>  
>  I´ve tried Ctrl-Alt-F1 but nothing happens, Ctrl-Alt-F2 shows
>  
>  IRQ handler tye mismatch for IRQ 0
>  current handler: timer
>  IRQ handler tye mismatch for IRQ 0
>  current handler: timer
>  IRQ handler tye mismatch for IRQ 0
>  current handler: timer
>  IRQ handler tye mismatch for IRQ 0
>  current handler: timer
>  hub 1-0:1.0: unable to enumerate USB device on port 1
>  
>  Any ideas?? I´m lost on this one... 
Looks its time to boot to run-level 1 and do some maintenance....

>  TIA
>  FC
>  PS: I don´t have a functioning CD/DVD reader on this system. I used at
>  the time of installation but not anymore... 

-- 
Kevin J. Cummings
kjchome(a)verizon.net
cummings(a)kjchome.homeip.net
cummings(a)kjc386.framingham.ma.us
Registered Linux User #1232 (http://counter.li.org)

[JB]

Fernando Cassia <fcassia <at> gmail.com> writes:

>  ... 
>  The issue is that I get the normal Fedora boot screen , then the
>  hourglass, then the log-in screen,then I get
>  
>  Oops, sealert hit an error!
>  
>  Traceback (most recent call last):
>  ... 
Stop boot loader, edit kernel boot line, and disable selinux for now.
   kernel /boot/vmlinuz-... ... selinux=0

This may get you thru ...

JB

[Fernando Cassia]

On Sun, Feb 27, 2011 at 7:12 PM, JB <jb.1234abcd(a)gmail.com&gt; wrote:
>  Stop boot loader, edit kernel boot line, and disable selinux for
> now.
>    kernel /boot/vmlinuz-... ... selinux=0
>
>  This may get you thru ...
>
>  JB 
Thanks, will try that.

Confirms my suspicion: SELinux´ best feature is to disable it. ;) (JOKE, JOKE)
FC

[Fernando Cassia]

On Sun, Feb 27, 2011 at 7:12 PM, JB <jb.1234abcd(a)gmail.com&gt; wrote:
>  Fernando Cassia <fcassia <at> gmail.com> writes:
>
> > ...
> > The issue is that I get the normal Fedora boot screen , then the
> > hourglass, then the log-in screen,then I get
> >
> > Oops, sealert hit an error!
> >
> > Traceback (most recent call last):
> > ...
>
>  Stop boot loader, edit kernel boot line, and disable selinux for now.
>    kernel /boot/vmlinuz-... ... selinux=0 
could it be possible that I´m still getting the same error, despite
pressing "a" (append) on the grub screen and appending selinux=0 at
the end of the line?

Scratching my head...

FC

[compdoc]

As Kevin J. Cummings said, you want to boot to runlevel 1, which is text mode. 
Boot to the command line, in other words.

Google:  fedora 12 runlevel 1

And if you are able to log in as root, you need to look at all of your 
partitions and see which one is full. If you aren't sure what to delete, ask 
the mailing list.

Don't mess with selinux.

[JB]

Fernando Cassia <fcassia <at> gmail.com> writes:

>  
>  On Sun, Feb 27, 2011 at 7:12 PM, JB <jb.1234abcd <at> gmail.com> wrote:
>  > Fernando Cassia <fcassia <at> gmail.com> writes:
>  >
>  >> ...
>  >> The issue is that I get the normal Fedora boot screen , then the
>  >> hourglass, then the log-in screen,then I get
>  >>
>  >> Oops, sealert hit an error!
>  >>
>  >> Traceback (most recent call last):
>  >> ...
>  >
>  > Stop boot loader, edit kernel boot line, and disable selinux for now.
>  >   kernel /boot/vmlinuz-... ... selinux=0
>  
>  could it be possible that I´m still getting the same error, despite
>  pressing "a" (append) on the grub screen and appending selinux=0 at
>  the end of the line?
>  
>  Scratching my head...
>  
>  FC 
Kernel boot parameters:
selinux=0            - disable SELinux entirely
enforcing=0          - disable SELinux enforcement (allow all)

Try appending both, perhaps one will stick ...

Be sure you append to kernel boot line that will actually boot your system.

You may also append number 1 (as suggested by other poster) to boot into
single user mode.

So it would be:
selinux=0 enforcing=0 1

Try it and report back.

JB

[Fernando Cassia]

On Sun, Feb 27, 2011 at 7:17 PM, Fernando Cassia <fcassia(a)gmail.com&gt; wrote:
>  Thanks, will try that.
>
>  Confirms my suspicion: SELinux´ best feature is to disable it. ;) (JOKE, JOKE)
>  FC 
Ok, runlevel 1 I went and erased some VirtualBox VM vdis taking a
couple gigs. Now space free is 98% according to "df". I hope SELinux
stops panicking now.

BTW: Is it "working as designed" to have an OS become comatose and
unable to boot when disk space is at 100% on the boot volume?.

Just sayin´... ;)

FC

[Joe Zeff]

On 02/27/2011 02:43 PM, compdoc wrote:
>  As Kevin J. Cummings said, you want to boot to runlevel 1, which is
> text mode.
>  Boot to the command line, in other words. 
Wrong.  Runlevel 1 is single-user mode.  You aren't connected to the net 
by default and only the root partition (and, if needed, /boot) mounted. 
  It's used for major system repairs when the system won't properly 
boot.  A text only boot, with everything mounted is runlevel 3.

[compdoc]

>  Wrong.  Runlevel 1 is single-user mode. 
Naw, I wasn't wrong - I was just explaining in a way anyone can understand.

[Fernando Cassia]

Please don´t fight, don´t need to. The problem was solved, thanks to you all.

A little manners goes a long way. No need to point fingers and accuse
others of being "wrong". :)

FC

On Sun, Feb 27, 2011 at 8:25 PM, compdoc <compdoc(a)hotrodpc.com&gt; wrote:
> > Wrong.  Runlevel 1 is single-user mode.
>
>  Naw, I wasn't wrong - I was just explaining in a way anyone can understand.

[JB]

Fernando Cassia <fcassia <at> gmail.com> writes:

>  ... 
>  Ok, runlevel 1 I went and erased some VirtualBox VM vdis taking a
>  couple gigs. Now space free is 98% according to "df". I hope SELinux
>  stops panicking now.
>  
>  BTW: Is it "working as designed" to have an OS become comatose and
>  unable to boot when disk space is at 100% on the boot volume?.
>  
>  Just sayin´... ;)
>  
>  FC 
Well, you may check your partitions and make some adjustment.
$ man tune2fs
$ tune2fs -l /dev/???
...
Block count:              3662812   <------- example       
Reserved block count:     36620     <------- example 1%      
...
# tune2fs -m 10 /dev/???            <--------example 10%

JB