[Build] Aggressively increase gradle based Test Cluster startup timeout (elastic#135672)

We see plenty of errors due to a cluster not starting up. This is likely a combination
of parallelism, memory pressure and to little timeout.

This is aggressively increasing the timeout from 40 to 120s while we will
monitor if the overall flakyness in ci is reduced

Author: breskeby

build-tools/src/main/java/org/elasticsearch/gradle/testclusters/ElasticsearchCluster.java

@@ -65,7 +65,7 @@
 public class ElasticsearchCluster implements TestClusterConfiguration, Named {
 
     private static final Logger LOGGER = Logging.getLogger(ElasticsearchNode.class);
-    private static final int CLUSTER_UP_TIMEOUT = 40;
+    private static final int CLUSTER_UP_TIMEOUT = 120;
     private static final TimeUnit CLUSTER_UP_TIMEOUT_UNIT = TimeUnit.SECONDS;
 
     private final AtomicBoolean configurationFrozen = new AtomicBoolean(false);