remove the nanny

we finally graduated high school and can make our own ramen

we no longer need this since fn appears to have no concept of canceling tasks
through an api we need to watch, and the context is plumbed if the request is
canceled. since tasks are short, we may never need to do cancellation of
running tasks like we had with iron worker. this was an added docker call
that's unnecessary since we are doing force removal of the container at the
end anyway.

api/runner/drivers/docker/docker_client.go

@@ -37,7 +37,6 @@ type dockerClient interface {
 	PullImage(opts docker.PullImageOptions, auth docker.AuthConfiguration) error
 	InspectImage(name string) (*docker.Image, error)
 	InspectContainer(id string) (*docker.Container, error)
-	StopContainer(id string, timeout uint) error
 	Stats(opts docker.StatsOptions) error
 }
 
@@ -281,16 +280,6 @@ func (d *dockerWrap) InspectContainer(id string) (c *docker.Container, err error
 	return c, err
 }
 
-func (d *dockerWrap) StopContainer(id string, timeout uint) (err error) {
-	ctx, cancel := context.WithTimeout(context.Background(), retryTimeout)
-	defer cancel()
-	err = d.retry(ctx, func() error {
-		err = d.docker.StopContainer(id, timeout)
-		return err
-	})
-	return filterNotRunning(ctx, filterNoSuchContainer(ctx, err))
-}
-
 func (d *dockerWrap) Stats(opts docker.StatsOptions) (err error) {
 	// we can't retry this one this way since the callee closes the
 	// stats chan, need a fancier retry mechanism where we can swap out