fix: delete cluster should not freeze while waiting for nodegroups …

…to be deleted

THis was due to a send/receive race-condition on an unbuffered golang channel. The waiter should run in a goroutine. Otherwise the waiter is unable to send to the errs channel, because there is no receiver yet.

pkg/cfn/manager/api.go

@@ -247,7 +247,7 @@ func (c *StackCollection) WaitDeleteStackTask(name string, errs chan error) erro
 
 	logger.Info("waiting for stack %q to get deleted", *i.StackName)
 
-	c.waitUntilStackIsDeleted(i, errs)
+	go c.waitUntilStackIsDeleted(i, errs)
 
 	return nil
 }