The following command sequence is ran: init_pid=$(lxc-info -n ${CONTAINER} --pid | awk '{ print $2 }') kill -PWR ${init_pid} <wait for pid to die> lxc-stop -n ${CONTAINER} However lxc-stop will complain the container is no longer started, because it has been killed before. Followed of course by the service being stuck. Would be probably be fixed if https://530896.bugs.gentoo.org/ is fixed. Reproducible: Always Steps to Reproduce: 1. Create a very fast LXC. 2. /etc/init.d/lxc.myfast start 3. /etc/init.d/lxc.myfast stop
I find the existing shutdown code rather obscure. Does that patch work for you? diff --git a/app-emulation/lxc/files/lxc.initd.3 b/app-emulation/lxc/files/lxc.initd.3 index c9ef999..b9d8c14 100644 --- a/app-emulation/lxc/files/lxc.initd.3 +++ b/app-emulation/lxc/files/lxc.initd.3 @@ -111,29 +111,8 @@ stop() { return 0 fi - init_pid=$(lxc-info -n ${CONTAINER} --pid | awk '{ print $2 }') - - if [ -z "${init_pid}" ]; then - ewarn "${CONTAINER} doesn't seem to be running." - return 0 - fi - - ebegin "Shutting down system in ${CONTAINER}" - kill -PWR ${init_pid} - eend $? - - TIMEOUT=${TIMEOUT:-30} - i=0 - while [ -n "$(pgrep -P ${init_pid})" -a $i -lt ${TIMEOUT} ]; do - sleep 1 - i=$(expr $i + 1) - done - - if [ -n "${missingprocs}" ]; then - ewarn "Something failed to properly shut down in ${CONTAINER}" - fi - + # 10s should be enough to shut everything down ebegin "Stopping ${CONTAINER}" - lxc-stop -n ${CONTAINER} + lxc-stop -t 30 -n ${CONTAINER} eend $?
I committed that fix
