test: fix race in simple/test-cluster-master-error

Said test checks that the workers shut down when the master errors but it failed
intermittently. Insert a small delay before doing the 'is dead?' check to give
the workers a chance to shut down.

test/simple/test-cluster-master-error.js

@@ -112,16 +112,21 @@ if (cluster.isWorker) {
     // Check that the cluster died accidently
     existMaster = (code === 1);
 
-    // When master is dead all workers should be dead to
-    var alive = false;
-    workers.forEach(function(pid) {
-      if (isAlive(pid)) {
-        alive = true;
-      }
-    });
+    // Give the workers time to shut down
+    setTimeout(checkWorkers, 200);
 
-    // If a worker was alive this did not act as expected
-    existWorker = !alive;
+    function checkWorkers() {
+      // When master is dead all workers should be dead to
+      var alive = false;
+      workers.forEach(function(pid) {
+        if (isAlive(pid)) {
+          alive = true;
+        }
+      });
+
+      // If a worker was alive this did not act as expected
+      existWorker = !alive;
+    }
   });
 
   process.once('exit', function() {