Hi,<div><br></div><div>I&#39;m trying to manage LXC instances through OpenStack, which use libvirt as a virtualization driver layer. After launching LXC instance, I simply could not attach to the console.</div><div><br></div>

<blockquote style="margin:0 0 0 40px;border:none;padding:0px"><div><div>virsh # list</div></div>
<div><div> Id Name                 State</div></div><div><div>----------------------------------</div></div><div><div>14366 instance-00000078    running</div></div><div><div><br></div></div><div><div>virsh # console 14366</div>


</div><div><div>Connected to domain instance-00000078</div></div><div><div>Escape character is ^]</div></div></blockquote><div><br></div><div>And it keeps stuck here.</div><div><br></div><div>Initially I thought it could be a cgroup device config issue, and tried to &quot;echo ‘c 5:1 rwm&#39; &gt;/cgroup/devices/libvirt/lxc/instance-00000078/devices.allow“, to add /dev/console access right to the container. But it does not work, console stilll not working.</div>

<div><br></div><div>Before the cgroup tweaking, devices.list are as follows</div><div><br></div><blockquote style="margin:0 0 0 40px;border:none;padding:0px"><div>[root@localhost libvirt]# cat /cgroup/devices/libvirt/lxc/instance-0000007a/devices.list</div>

<div><br></div><div><div>c 1:3 rwm</div></div><div><div>c 1:5 rwm</div></div><div><div>c 1:7 rwm</div></div><div><div>c 1:8 rwm</div></div><div><div>c 1:9 rwm</div></div><div><div>c 5:0 rwm</div></div><div><div>c 5:2 rwm</div>

</div><div><div>c 136:* rwm</div></div></blockquote>
<div><br></div><div>I&#39;m not quite sure what kind of issue could possibly cause this console access issue. Cause I&#39;m now using two different guest OS templates on two different host OSes. Following are some symptoms I observed.</div>

<div><blockquote style="margin:0 0 0 40px;border:none;padding:0px"><div><br></div><div>* (g) for container guest OS</div><div>* (h) for host OS</div><div>* NOTE: Ubuntu guest os are generated by lxc-create -t ubuntu, which is an Ubuntu 12.04 based lxc template.</div>

<div>             CentOS guest os are handcrafted by myself, taking reference of <a href="http://wiki.1tux.org/wiki/Lxc/Installation/Guest/Centos/6">http://wiki.1tux.org/wiki/Lxc/Installation/Guest/Centos/6</a></div><div>

<ol><li>Ubuntu (g) launched by lxc-start on Ubuntu 12.04 host, console works</li><li>Ubuntu (g) launched by OpenStack / libvirt on Ubuntu 12.04 host, console works</li><li>Ubuntu (g) launched by OpenStack / libvirt on CentOS 6.3 host, console NOT works</li>

<li>CentOS (g) launched by lxc-start on Ubuntu 12.04 host, console works</li><li>CentOS (g) launched by OpenStack / libvirt on Ubuntu12.04 host, console NOT works</li><li>CentOS (g) launched by OpenStack / libvirt on CentOS 6.3 host, console NOT works</li>

</ol></div></blockquote></div><div>So somehow I feel it could be something related to LXC template rootfs, but Host OS do make a difference (2 vs. 3). lxc-start / libvirt also makes a difference (4 vs. 5), but that might be caused by cgroup setting.</div>

<div><br></div><div>Could someone shed me some clues to further digging this issue? Thanks!  </div><div><br></div><div>--<br>Qiu Yu<br>
</div>