NIFI-12221: This closes #7876. Be more lenient about which Disconnect…

…ion Codes we allow a node to be reconnected to a cluster vs. when we notify the node to disconnect again. Also updated the timeout for OffloadIT because it occasionally times ou out while running properly. Signed-off-by: Joseph Witt <joewitt@apache.org>
apache · Oct 13, 2023 · f4ae292 · f4ae292
1 parent 0eabbcd
commit f4ae292
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 5 deletions.
diff --git a/...rc/main/java/org/apache/nifi/cluster/coordination/heartbeat/AbstractHeartbeatMonitor.java b/...rc/main/java/org/apache/nifi/cluster/coordination/heartbeat/AbstractHeartbeatMonitor.java
@@ -254,20 +254,22 @@ private void processHeartbeat(final NodeHeartbeat heartbeat) {
                 case LACK_OF_HEARTBEAT:
                 case UNABLE_TO_COMMUNICATE:
                 case NOT_YET_CONNECTED:
-                case STARTUP_FAILURE: {
+                case MISMATCHED_FLOWS:
+                case MISSING_BUNDLE:
+                case NODE_SHUTDOWN:
+                case FAILED_TO_SERVICE_REQUEST:
+                case STARTUP_FAILURE:
                     clusterCoordinator.reportEvent(nodeId, Severity.INFO, "Received heartbeat from node previously "
                             + "disconnected due to " + disconnectionCode + ". Issuing reconnection request.");
 
                     clusterCoordinator.requestNodeConnect(nodeId, null);
                     break;
-                }
-                default: {
+                default:
                     // disconnected nodes should not heartbeat, so we need to issue a disconnection request.
-                    logger.info("Ignoring received heartbeat from disconnected node " + nodeId + ".  Issuing disconnection request.");
+                    logger.info("Ignoring received heartbeat from disconnected node {}. Node was disconnected due to [{}]. Issuing disconnection request.", nodeId, disconnectionCode);
                     clusterCoordinator.requestNodeDisconnect(nodeId, disconnectionCode, connectionStatus.getReason());
                     removeHeartbeat(nodeId);
                     break;
-                }
             }
 
             return;

diff --git a/...fi-system-test-suite/src/test/java/org/apache/nifi/tests/system/clustering/OffloadIT.java b/...fi-system-test-suite/src/test/java/org/apache/nifi/tests/system/clustering/OffloadIT.java
@@ -25,11 +25,13 @@
 import org.apache.nifi.web.api.entity.ConnectionEntity;
 import org.apache.nifi.web.api.entity.ProcessorEntity;
 import org.junit.jupiter.api.Test;
+import org.junit.jupiter.api.Timeout;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 import java.io.IOException;
 import java.util.Collections;
+import java.util.concurrent.TimeUnit;
 
 public class OffloadIT extends NiFiSystemIT {
     private static final Logger logger = LoggerFactory.getLogger(OffloadIT.class);
@@ -40,6 +42,9 @@ public NiFiInstanceFactory getInstanceFactory() {
     }
 
     @Test
+    @Timeout(value = 10, unit = TimeUnit.MINUTES)
+    // Test to ensure that node can be offloaded, reconnected, offloaded several times. This test typically takes only about 1-2 minutes
+    // but can occasionally take 5-6 minutes on Github Actions so we set the timeout to 10 minutes to allow for these occasions
     public void testOffload() throws InterruptedException, IOException, NiFiClientException {
         for (int i=0; i < 5; i++) {
             logger.info("Running iteration {}", i);