engine: graph: Prevent converged timeout of dirty res

Somewhere after the engine re-write we seem to have regressed and
converge early even if some resource is dirty. This adds an additional
timer so that we don't start the individual resource converged countdown
until our state is okay.

engine/graph/actions.go

@@ -119,6 +119,7 @@ func (obj *Engine) Process(vertex pgraph.Vertex) error {
 			for _, changed := range updated {
 				if changed { // at least one was updated
 					// invalidate cache, mark as dirty
+					obj.state[vertex].tuid.StopTimer()
 					obj.state[vertex].isStateOK = false
 					break
 				}
@@ -174,6 +175,7 @@ func (obj *Engine) Process(vertex pgraph.Vertex) error {
 
 	// if CheckApply ran without noop and without error, state should be good
 	if !noop && err == nil { // aka !noop || checkOK
+		obj.state[vertex].tuid.StartTimer()
 		obj.state[vertex].isStateOK = true // reset
 		if refresh {
 			obj.SetUpstreamRefresh(vertex, false) // refresh happened, clear the request
@@ -252,9 +254,11 @@ func (obj *Engine) Worker(vertex pgraph.Vertex) error {
 	defer close(obj.state[vertex].stopped) // done signal
 
 	obj.state[vertex].cuid = obj.Converger.Register()
+	obj.state[vertex].tuid = obj.Converger.Register()
 	// must wait for all users of the cuid to finish *before* we unregister!
 	// as a result, this defer happens *before* the below wait group Wait...
 	defer obj.state[vertex].cuid.Unregister()
+	defer obj.state[vertex].tuid.Unregister()
 
 	defer obj.state[vertex].wg.Wait() // this Worker is the last to exit!