Test: Write node ports file before starting tribe service (#24351)

The tribe service can take a while to initialize, depending on how many cluster it needs to connect to. This change moves writing the ports file used by tests to before the tribe service is started.

core/src/main/java/org/elasticsearch/common/settings/ClusterSettings.java

@@ -61,7 +61,6 @@ import org.elasticsearch.discovery.zen.ZenDiscovery;
 import org.elasticsearch.env.Environment;
 import org.elasticsearch.env.NodeEnvironment;
 import org.elasticsearch.gateway.GatewayService;
-import org.elasticsearch.gateway.PrimaryShardAllocator;
 import org.elasticsearch.http.HttpTransportSettings;
 import org.elasticsearch.index.IndexSettings;
 import org.elasticsearch.indices.IndexingMemoryController;
@@ -345,7 +344,7 @@ public final class ClusterSettings extends AbstractScopedSettings {
                     SearchService.DEFAULT_KEEPALIVE_SETTING,
                     SearchService.KEEPALIVE_INTERVAL_SETTING,
                     SearchService.LOW_LEVEL_CANCELLATION_SETTING,
-                    Node.WRITE_PORTS_FIELD_SETTING,
+                    Node.WRITE_PORTS_FILE_SETTING,
                     Node.NODE_NAME_SETTING,
                     Node.NODE_DATA_SETTING,
                     Node.NODE_MASTER_SETTING,

core/src/main/java/org/elasticsearch/node/Node.java

@@ -168,7 +168,7 @@ import static java.util.stream.Collectors.toList;
 public class Node implements Closeable {
 
 
-    public static final Setting<Boolean> WRITE_PORTS_FIELD_SETTING =
+    public static final Setting<Boolean> WRITE_PORTS_FILE_SETTING =
         Setting.boolSetting("node.portsfile", false, Property.NodeScope);
     public static final Setting<Boolean> NODE_DATA_SETTING = Setting.boolSetting("node.data", true, Property.NodeScope);
     public static final Setting<Boolean> NODE_MASTER_SETTING =
@@ -727,13 +727,7 @@ public class Node implements Closeable {
             injector.getInstance(HttpServerTransport.class).start();
         }
 
-        // start nodes now, after the http server, because it may take some time
+        if (WRITE_PORTS_FILE_SETTING.get(settings)) {
-        tribeService.startNodes();
-        // starts connecting to remote clusters if any cluster is configured
-        SearchTransportService searchTransportService = injector.getInstance(SearchTransportService.class);
-        searchTransportService.start();
-
-        if (WRITE_PORTS_FIELD_SETTING.get(settings)) {
             if (NetworkModule.HTTP_ENABLED.get(settings)) {
                 HttpServerTransport http = injector.getInstance(HttpServerTransport.class);
                 writePortsFile("http", http.boundAddress());
@@ -742,6 +736,12 @@ public class Node implements Closeable {
             writePortsFile("transport", transport.boundAddress());
         }
 
+        // start nodes now, after the http server, because it may take some time
+        tribeService.startNodes();
+        // starts connecting to remote clusters if any cluster is configured
+        SearchTransportService searchTransportService = injector.getInstance(SearchTransportService.class);
+        searchTransportService.start();
+
         logger.info("started");
 
         return this;

plugins/discovery-azure-classic/src/test/java/org/elasticsearch/discovery/azure/classic/AzureDiscoveryClusterFormationTests.java

@@ -109,7 +109,7 @@ public class AzureDiscoveryClusterFormationTests extends ESIntegTestCase {
             .put(DiscoveryModule.DISCOVERY_TYPE_SETTING.getKey(), AzureDiscoveryPlugin.AZURE)
             .put(Environment.PATH_LOGS_SETTING.getKey(), resolve)
             .put(TransportSettings.PORT.getKey(), 0)
-            .put(Node.WRITE_PORTS_FIELD_SETTING.getKey(), "true")
+            .put(Node.WRITE_PORTS_FILE_SETTING.getKey(), "true")
             .put(AzureComputeService.Management.ENDPOINT_SETTING.getKey(), "https://" + InetAddress.getLoopbackAddress().getHostAddress() +
                 ":" + httpsServer.getAddress().getPort())
             .put(Environment.PATH_CONF_SETTING.getKey(), keyStoreFile.getParent().toAbsolutePath())

qa/smoke-test-tribe-node/build.gradle

@@ -52,4 +52,14 @@ integTestCluster {
   setting 'tribe.two.http.enabled', 'true'
   setting 'tribe.two.http.port', '40250-40299'
   setting 'tribe.two.transport.tcp.port', '40250-40399'
+
+  waitCondition = { node, ant ->
+    File tmpFile = new File(node.cwd, 'wait.success')
+    // 5 nodes: tribe + clusterOne (1 node + tribe internal node) + clusterTwo (1 node + tribe internal node)
+    ant.get(src: "http://${node.httpUri()}/_cluster/health?wait_for_nodes=>=5&wait_for_status=yellow",
+            dest: tmpFile.toString(),
+            ignoreerrors: true,
+            retries: 10)
+    return tmpFile.exists()
+  }
 }