Kernel module loading is broken on 3.8.0-25?

[ku1ik]

I upgraded my kernel on Ubuntu 13.04 recently (just by apt-get upgrade) and restarted.
Since then when I try to enable firewall on a raring VM with ufw enable I get following:

[Errno 2] libkmod: ERROR ../libkmod/libkmod.c:554 kmod_search_moddep: could not open moddep file '/lib/modules/3.8.0-25-generic/modules.dep.bin'
ip6tables v1.4.12: can't initialize ip6tables table `filter': Table does not exist (do you need to insmod?)
Perhaps ip6tables or your kernel needs to be upgraded.

I checked /lib/modules/ directory and it's not there but that's the same raring image that worked before so the issue must be hidden somewhere else.

Full output (from vagrant up to the above error) here:
https://gist.github.com/sickill/5935636

[fgrehm]

@sickill funny that you mentioned that, my laptop just crashed for some weird reason and I was only able to boot again by rolling back to 3.8.0-19, I can't really reboot right now to dig into it but I'll keep an eye on this. tks for reporting!

[fgrehm]

@sickill quick update, been using 3.8.0-25 over the last week and things are working fine. were you able to dig into the issue? I was able to reproduce the problem with ufw you mentioned but I have no idea how to fix it =/ I'll give a shot at using an older kernel to check if things work fine when I have a chance

[ku1ik]

@fgrehm unfortunatelly I haven't found the cause of the problem.

[fgrehm]

Does booting with a different version work?
On Jul 13, 2013 4:14 PM, "Marcin Kulik" notifications@github.com wrote:

@fgrehm https://github.com/fgrehm unfortunatelly I haven't found the
cause of the problem.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/104#issuecomment-20924983
.

[ku1ik]

Unfortunatelly (but fortunately for me) I switched to Fedora recently and I don't have this old Ubuntu installation anymore to check on some different kernel.

[fgrehm]

@sickill No worries, I'll have a look at it from here. So you've been using vagrant-lxc on Fedora? If yes, did it require any manual intervetion? Would you mind adding some notes to this wiki entry? :)

[fgrehm]

@sickill oh, just noticed you've created #113, sorry about that :)

[fgrehm]

updates on this one: I've just upgraded to 3.8.0-26 and I still can't use ufw on vagrant-lxc containers. I'll try this out with vanilla lxc templates to check if it works.

[fgrehm]

@sickill I'm not sure what's going on here, I've added a link to this issue to our Troubleshooting section of the Wiki in case someone hits the same problem.

I'm up for fixing things from vagrant-lxc side if the problem is there :)

[fgrehm]

So I've hit this today while trying to run Docker inside a vagrant-lxc container. It used to work just fine on the past but I think they might have changed something at some point or maybe I messed up with something else :P

For now I just went ahead and shared /lib/modules with the guest container and Docker was able to work without issues. I'm not sure what are the implications of that I believe it shouldn't be a problem for dev environments. If things end up working fine I'll just add an entry to our troubleshooting wiki page

[fgrehm]

Turns out that my issue was actually related to not providing lxc.customize 'aa_profile', 'unconfined'

[abevoelker]

I was getting a similar (but different) error related to kernel modules when trying to install the iptables-persistent package in a container running within Ubuntu 12.04 w/ the 3.8.0-33 kernel:

vagrant@precise-base:~$ sudo apt-get -y install iptables-persistent
Reading package lists... Done
Building dependency tree       
Reading state information... Done
iptables-persistent is already the newest version.
0 upgraded, 0 newly installed, 0 to remove and 20 not upgraded.
1 not fully installed or removed.
After this operation, 0 B of additional disk space will be used.
Setting up iptables-persistent (0.5.3ubuntu2) ...
FATAL: Could not load /lib/modules/3.8.0-33-generic/modules.dep: No such file or directory
dpkg: error processing iptables-persistent (--configure):
 subprocess installed post-installation script returned error exit status 1
Errors were encountered while processing:
 iptables-persistent
E: Sub-process /usr/bin/dpkg returned an error code (1)

Turns out the container's /lib/modules directory is empty, I'm assuming for security reasons:

vagrant@precise-base:~$ sudo find /lib/modules
/lib/modules

Neither installing iptables-persistent on the host machine nor adding lxc.customize 'aa_profile', 'unconfined' fixed the issue for me. I'm sure it's really dangerous, but the fix I'm using for now is to just share /lib/modules on the host with the container:

config.vm.synced_folder "/lib/modules", "/lib/modules"

[iamjvn]

It seems this recent change has broken the synced_folder "/lib/modules" workaround.

lxc-start 20170622002807.856 ERROR    lxc_utils - utils.c:open_without_symlink:1682 - Too many levels of symbolic links - lib in /usr/lib/x86_64-linux-gnu/lxc/lib/modules was a symbolic link!
lxc-start 20170622002807.857 ERROR    lxc_conf - conf.c:mount_entry:1708 - Too many levels of symbolic links - failed to mount '/lib/modules' on '/usr/lib/x86_64-linux-gnu/lxc/lib/modules'