[Lxc-users] [PATCH 9/9] lxc-destroy: wait until the container is stopped

[Serge Hallyn]

Quoting Daniel Lezcano (daniel.lezcano at free.fr):
> On 04/26/2012 07:09 AM, Serge Hallyn wrote:
> >From: Serge Hallyn<serge.hallyn at ubuntu.com>
> >
> >Author: Timothy Chen<tnachen at gmail.com>
> >Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/980902
> >
> >Signed-off-by: Serge Hallyn<serge.hallyn at ubuntu.com>
> >Cc: Timothy Chen<tnachen at gmail.com>
> >---
> >  src/lxc/lxc-destroy.in |    1 +
> >  1 file changed, 1 insertion(+)
> >
> >diff --git a/src/lxc/lxc-destroy.in b/src/lxc/lxc-destroy.in
> >index b0f2da5..17fa6d6 100644
> >--- a/src/lxc/lxc-destroy.in
> >+++ b/src/lxc/lxc-destroy.in
> >@@ -87,6 +87,7 @@ lxc-info -n $lxc_name 2>/dev/null | grep -q RUNNING
> >  if [ $? -eq 0 ]; then
> >  	if [ $force -eq 1 ]; then
> >  		lxc-stop -n $lxc_name
> >+		lxc-wait -n $lxc_name -s STOPPED
> >  	else
> >  		echo "Container $lxc_name is running, aborting the deletion."
> >  		exit 1
> 
> I don't get why this is needed, lxc-stop is synchronous. When the
> lxc-stop command exits, we have the guarantee the container has
> stopped. If it is not the case, that means there is a problem
> somewhere else.

Some of the filesystems may not yet have been auto-umounted when the
namespaces were cleaned up, causing the subsequent rm -rf $rootfs to
fail.

-serge