As discussed in IRC, to make sure child processes are cleaned up before a worker receives SIGKILL, when receiving the finish event from Mojo::IOLoop, kill child processes after a timeout based on the default graceful_timeout. Allow configuration of timeout and document relation to the graceful_timeout.
I know that ForkCall is now deprecated, but is there away to pull off the timeout using ForkCall's successor? I've searched around and can't turn up details or related examples.
As discussed in IRC, to make sure child processes are cleaned up before a worker receives SIGKILL, when receiving the finish event from Mojo::IOLoop, kill child processes after a timeout based on the default graceful_timeout. Allow configuration of timeout and document relation to the graceful_timeout.