[Lxc-users] uptodate Ubuntu Lucid guests

Papp Tamás tompos at martos.bme.hu
Sat Sep 4 22:58:28 UTC 2010


hi All,

1. I have some more problem. I guest a hard lockup. I really don't know, 
why. There was no high load or any fs activity. I just run 
/etc/init.d/mailman start inside the VM and got an oops message on the 
console. Unfortunately after the reboot the logs were empty. Sure I 
cannot reproduce it, at least I hope.

Also still I could not reproduce the error related to lxc-stop, reported 
by me earlier.

The host is an Ubuntu Lucid 10.04 x86_64 with lxc 0.7.2 (another machine 
than in the bugreport some hours ago).

2. Some days ago there was an update to the package udev. In the 
postinst script there is section:

create_devices()
{
   # in a vserver environment, mknod will fail; cf. LP: #144685
   if grep -q ^VxID /proc/self/status; then
   return
   fi

What is the right grep pattern so it matches for lxc? In other words, 
how can I be sure, I'm in an lxc container?

3. In the last weeks there were some updates to ubuntu. Unfortunately I 
didn't restarted my containers after the updates, so I've just realized 
now, that the containers cannot boot properly.
I just cannot figure out, why.

I don't see, what the proper configuration should be, or what is 
depending on what. There is a fully working unupdated configuration, an 
updated, but not working at all, and an updated half-working 
configurations (sshd is starting other services not).
I made some comparisons but looks similar or even the same to me. Could 
you give me some help on this, how could I debug it?

4. if I start the container without the -d flag, so the command is 
lxc-start -n name, should I see the console messages? Now I don't see them.
Is it possible to make it work like xen does it with the command 'xm 
create foo.cfg -c'?

Thank you,

tamas





More information about the lxc-users mailing list