[gentoo-server] can't boot anymore!

[gentoo-server] can't boot anymore!

[Davide Cittaro]

Hi there, I think I've messed up something in init scripts and  
probably I'm missing something... btw I have some machines, different  
hardware, and they all can't boot.
When udev is used (as specified in /etc/conf.d/rc) it crashes and  
disks can't be created, so I have only an emergency shell...
If I specify static devices (also in /etc/conf.d/rc) I have fsck  
working on disks, but later on it freezes on "Caching services  
dependendecies"... If I interrupt it says that /var/lib/init.d/ 
depcache doesn't exists and /sbin/rc fails with a segfault on "chgrp  
utmp /var/run/utmp"...
I really can't get rid of this... Since I've never touched udev rules  
or init scripts I guess it somehow has to do with some broken ebuild...

Any hint?

Davide Cittaro
davide.cittaro@ifom-ieo-campus.it

--
Give a man a fish, and he eats for a day. Teach a man to phish, and if  
he gets caught he'll be eating that fish through a straw



-- 
gentoo-server@lists.gentoo.org mailing list

Re: [gentoo-server] can't boot anymore!

[Randy Barlow]

Davide Cittaro wrote:
> Hi there, I think I've messed up something in init scripts and probably 
> I'm missing something... btw I have some machines, different hardware, 
> and they all can't boot.
> When udev is used (as specified in /etc/conf.d/rc) it crashes and disks 
> can't be created, so I have only an emergency shell...
> If I specify static devices (also in /etc/conf.d/rc) I have fsck working 
> on disks, but later on it freezes on "Caching services 
> dependendecies"... If I interrupt it says that /var/lib/init.d/depcache 
> doesn't exists and /sbin/rc fails with a segfault on "chgrp utmp 
> /var/run/utmp"...
> I really can't get rid of this... Since I've never touched udev rules or 
> init scripts I guess it somehow has to do with some broken ebuild...
> 
> Any hint?

What did you do?

-- 
Randy Barlow
http://electronsweatshop.com
-- 
gentoo-server@lists.gentoo.org mailing list

Re: [gentoo-server] can't boot anymore!

[Davide Cittaro]

[-- Attachment #1: Type: text/plain, Size: 478 bytes --]


On Apr 10, 2008, at 5:13 PM, Randy Barlow wrote:
>
> What did you do?


If I only knew :-(
The fact is that I've never rebooted in a year but, in the meanwhile,  
I've kept system up-to-date by regular emerge syncing and system  
upgrade... evidently I've emerged (or unmerged) something wrong...

d


/*
Davide Cittaro

Cogentech - Consortium for Genomic Technologies
via adamello, 16
20139 Milano
Italy

tel.: +39(02)574303007
e-mail: davide.cittaro@ifom-ieo-campus.it
*/




[-- Attachment #2: Type: text/html, Size: 2771 bytes --]

Re: [gentoo-server] can't boot anymore!

[Kerin Millar]

On 10/04/2008, Davide Cittaro <davide.cittaro@ifom-ieo-campus.it> wrote:
> Hi there, I think I've messed up something in init scripts and probably I'm
> missing something... btw I have some machines, different hardware, and they
> all can't boot.
>  When udev is used (as specified in /etc/conf.d/rc) it crashes and disks
> can't be created, so I have only an emergency shell...
>  If I specify static devices (also in /etc/conf.d/rc) I have fsck working on
> disks, but later on it freezes on "Caching services dependendecies"... If I
> interrupt it says that /var/lib/init.d/depcache doesn't exists and /sbin/rc
> fails with a segfault on "chgrp utmp /var/run/utmp"...
>  I really can't get rid of this... Since I've never touched udev rules or
> init scripts I guess it somehow has to do with some broken ebuild...

I've seen a similar issue mentioned in the forums before. Ensure that
/etc/init.d only contains valid scripts (no incomplete merges and the
like) and also that there are no orphaned files (qfile -o
/etc/init.d/*). Then try touch /etc/init.d/*; /sbin/depscan.sh
--update.

Regards,

--Kerin
-- 
gentoo-server@lists.gentoo.org mailing list

Re: [gentoo-server] can't boot anymore!

[Davide Cittaro]

[-- Attachment #1: Type: text/plain, Size: 539 bytes --]


On Apr 10, 2008, at 5:32 PM, Kerin Millar wrote:
>
> I've seen a similar issue mentioned in the forums before. Ensure that
> /etc/init.d only contains valid scripts (no incomplete merges and the
> like) and also that there are no orphaned files (qfile -o
> /etc/init.d/*). Then try touch /etc/init.d/*; /sbin/depscan.sh
> --update.
>

Thanks, I'll try that!

d


/*
Davide Cittaro

Cogentech - Consortium for Genomic Technologies
via adamello, 16
20139 Milano
Italy

tel.: +39(02)574303007
e-mail: davide.cittaro@ifom-ieo-campus.it
*/




[-- Attachment #2: Type: text/html, Size: 2856 bytes --]

Re: [gentoo-server] can't boot anymore!

[Andreas Tasch]

Davide Cittaro schrieb:
> 
> On Apr 10, 2008, at 5:13 PM, Randy Barlow wrote:
>>
>> What did you do?
> 
> If I only knew :-(
> The fact is that I've never rebooted in a year but, in the meanwhile, 
> I've kept system up-to-date by regular emerge syncing and system 
> upgrade... evidently I've emerged (or unmerged) something wrong...
> 
> d
> 
> 
> /*
> Davide Cittaro
> 
> Cogentech - Consortium for Genomic Technologies
> via adamello, 16
> 20139 Milano
> Italy
> 
> tel.: +39(02)574303007
> e-mail: davide.cittaro@ifom-ieo-campus.it 
> <mailto:davide.cittaro@ifom-ieo-campus.it>
> */

Hi Davide,

sounds like the problem I had some time ago. I did an gcc upgrade but 
did not do an emerge -e system after that. Something was broken after 
that and the system did not boot anymore after I had to reboot some 
months later (it was a remote machine :(). What solved my problem: boot 
with a live cd or rescue image via pxe and chroot in your system. Do an 
source /etc/profile and env-update (afairc). Then re-emerge gcc glibc 
binutils and baselayout and reboot. Do a backup before trying this - you 
never know :)

Good Luck,
Andy

-- 
gentoo-server@lists.gentoo.org mailing list

Re: [gentoo-server] can't boot anymore!

[Bastiaan Visser]

Andreas Tasch wrote:
> Davide Cittaro schrieb:
>>
>> On Apr 10, 2008, at 5:13 PM, Randy Barlow wrote:
>>>
>>> What did you do?
>>
>> If I only knew :-(
>> The fact is that I've never rebooted in a year but, in the meanwhile, 
>> I've kept system up-to-date by regular emerge syncing and system 
>> upgrade... evidently I've emerged (or unmerged) something wrong...
>>
>> d
>>
>>
>> /*
>> Davide Cittaro
>>
>> Cogentech - Consortium for Genomic Technologies
>> via adamello, 16
>> 20139 Milano
>> Italy
>>
>> tel.: +39(02)574303007
>> e-mail: davide.cittaro@ifom-ieo-campus.it 
>> <mailto:davide.cittaro@ifom-ieo-campus.it>
>> */
>
> Hi Davide,
>
> sounds like the problem I had some time ago. I did an gcc upgrade but 
> did not do an emerge -e system after that. Something was broken after 
> that and the system did not boot anymore after I had to reboot some 
> months later (it was a remote machine :(). What solved my problem: 
> boot with a live cd or rescue image via pxe and chroot in your system. 
> Do an source /etc/profile and env-update (afairc). Then re-emerge gcc 
> glibc binutils and baselayout and reboot. Do a backup before trying 
> this - you never know :)
>
> Good Luck,
> Andy
>
A revdep-rebuild might be enough then, saves you  a day or two compiling.
-- 
gentoo-server@lists.gentoo.org mailing list

Re: [gentoo-server] can't boot anymore!

[Davide Cittaro]

[-- Attachment #1: Type: text/plain, Size: 1014 bytes --]

Hi all and thanks for many answers. Apparently none of them solve my  
problem, I still get stuck at boot, before entering in runlevel 3.
I've also reinstalled a system from scratch and it doesn't work. I  
have the hint that everything is related with hostname resolving. As  
an example, when I chroot from a livecd I cannot run any command  
(except for cat, touch and few others) so I can't even ls directories.  
If I delete the /etc/hosts file I get everything working.
Yesterday, when I've was investigating, I had udevd that crashes.  
Running with strace I see that it segfaults after reading /etc/hosts.  
deleting it makes it crash later on and, with strace, I can see a  
continous attempt to resolve the hostname by address... Can this be  
the reason why I have all these problems? Did anybody have a similar  
problem?

d


/*
Davide Cittaro

Cogentech - Consortium for Genomic Technologies
via adamello, 16
20139 Milano
Italy

tel.: +39(02)574303007
e-mail: davide.cittaro@ifom-ieo-campus.it
*/




[-- Attachment #2: Type: text/html, Size: 3144 bytes --]

Re: [gentoo-server] can't boot anymore!

[Kalin KOZHUHAROV]

Hi Davide,

Davide Cittaro wrote:
> Yesterday, when I've was investigating, I had udevd that crashes. 
> Running with strace I see that it segfaults after reading /etc/hosts. 
> deleting it makes it crash later on and, with strace, I can see a 
> continous attempt to resolve the hostname by address...
Well, then can you post attach /etc/hosts in question?
You might have hit 1 in 1M bug :-)

> Can this be the reason why I have all these problems?
> Did anybody have a similar problem?
If you state that you installed from scratch, it may well be connected.
Think if what specific you did, anything apart form the Gentoo Handbook?

Cheers,

Kalin.

-- 
|[ ~~~~~~~~~~~~~~~~~~~~~~ ]|
+-> http://ThinRope.net/ <-+
|[ ______________________ ]|

-- 
gentoo-server@lists.gentoo.org mailing list

[gentoo-server] Fixed! [was: can't boot anymore!]

[Davide Cittaro]

[-- Attachment #1: Type: text/plain, Size: 603 bytes --]

Ok, I've fixed the thing...
I was right that something in /etc/something was wrong.. indeed /etc/ 
nsswitch.conf had this line:

groups	ldap files

in place of

groups	files ldap

The line should be ok but for nsswitch.ldap !!!
As the system booted could not find correct group (i.e. utmp) and  
everything got stuck. Also chroot didn't work for the same reason.
  BTW, thanks to all, I will pay double attention next time!

d

/*
Davide Cittaro

Cogentech - Consortium for Genomic Technologies
via adamello, 16
20139 Milano
Italy

tel.: +39(02)574303007
e-mail: davide.cittaro@ifom-ieo-campus.it
*/




[-- Attachment #2: Type: text/html, Size: 2863 bytes --]