On Wed, Mar 27, 2013 at 3:38 PM, Ben Pfaff <b...@nicira.com> wrote:
> On Wed, Mar 27, 2013 at 02:30:05PM -0700, Gurucharan Shetty wrote:
>> Currently, when we stop a daemon, we first send it SIGTERM.
>> If SIGTERM did not work within ~5 seconds, we send a SIGKILL.
>> After sending SIGKILL, we wait only for 4 seconds, before giving
>> up.
>>
>> If the system is exteremely busy, there is a chance that a
>> process is not killed by the kernel within 4 seconds. In such
>> a case, when we try to start the daemon immediately, we see that
>> the pid inside the pid-file is valid and assume that the daemon
>> is still running. This leaves us in a state, where the daemon is
>> actually not running.
>>
>> This patch increases the time waiting for the kernel to kill the
>> process to 60 seconds.
>>
>> Bug #15404.
>> Signed-off-by: Gurucharan Shetty <gshe...@nicira.com>
>
> I see why you changed the FAIL case, but I think that it might instead
> be better to do something like the following, to avoid duplicating the
> pid_exists call in two places:
>
> diff --git a/utilities/ovs-lib.in b/utilities/ovs-lib.in
> index d010abf..b44ab37 100644
> --- a/utilities/ovs-lib.in
> +++ b/utilities/ovs-lib.in
> @@ -173,6 +173,9 @@ stop_daemon () {
> if test -e "$rundir/$1.pid"; then
> if pid=`cat "$rundir/$1.pid"`; then
> for action in TERM .1 .25 .65 1 1 1 1 KILL 1 1 1 1 FAIL; do
> + if pid_exists $pid >/dev/null 2>&1; then :; else
> + return 0
> + fi
Yes, this is better. I will make the change and commit this to master and 1.10
> case $action in
> TERM)
> action "Killing $1 ($pid)" kill $pid
> @@ -185,11 +188,7 @@ stop_daemon () {
> return 1
> ;;
> *)
> - if pid_exists $pid >/dev/null 2>&1; then
> - sleep $action
> - else
> - return 0
> - fi
> + sleep $action
> ;;
> esac
> done
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
