(ignite) branch master updated: IGNITE-28830 Flaky CacheGroupsMetricsRebalanceTest#testRebalanceProgressUnderLoad: EVT_CACHE_REBALANCE_STOPPED listener is registered after the joining node starts rebalancing (#13291)

[av]

This is an automated email from the ASF dual-hosted git repository.

anton-vinogradov pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/ignite.git


The following commit(s) were added to refs/heads/master by this push:
     new 2a5af087998 IGNITE-28830 Flaky 
CacheGroupsMetricsRebalanceTest#testRebalanceProgressUnderLoad: 
EVT_CACHE_REBALANCE_STOPPED listener is registered after the joining node 
starts rebalancing (#13291)
2a5af087998 is described below

commit 2a5af087998d57d58c7e7b33ad0ec36a97bbc16c
Author: Anton Vinogradov <a...@apache.org>
AuthorDate: Mon Jun 29 13:41:12 2026 +0300

    IGNITE-28830 Flaky 
CacheGroupsMetricsRebalanceTest#testRebalanceProgressUnderLoad: 
EVT_CACHE_REBALANCE_STOPPED listener is registered after the joining node 
starts rebalancing (#13291)
    
    ## Description
    
    `testRebalanceProgressUnderLoad()` registered an
    `EVT_CACHE_REBALANCE_STOPPED` listener via
    `ignite.events().localListen(...)` **after** `startGrid(4)`. The joining
    node begins rebalancing during `start()`, so the event could fire before
    the listener was registered. The event was then missed, the
    `CountDownLatch` was never counted down, and the unbounded
    `latch.await()` hung until the 600s test timeout (reproduced in ~7% of
    runs; a thread dump shows the test runner parked in `latch.await()`).
    
    ## Fix
    
    Register the listener through node configuration
    (`IgniteConfiguration.setLocalEventListeners`) so it is active before
    the node joins and starts rebalancing. Both `events().localListen()` and
    `setLocalEventListeners` funnel into
    `GridEventStorageManager.addLocalEventListener → addEventListener →
    registerListener`, and neither enables the event type — so delivery is
    identical; the listener is simply active earlier (during node
    `start()`), eliminating the race.
    
    ## Verification
    
    - Compiles cleanly.
    - `testRebalanceProgressUnderLoad` ran green across repeated local runs
    (~0.5s each), no hang.
    
    JIRA: https://issues.apache.org/jira/browse/IGNITE-28830
    
    Co-authored-by: Claude Opus 4.8 <nore...@anthropic.com>
---
 .../cache/CacheGroupsMetricsRebalanceTest.java     | 27 +++++++++++-----------
 1 file changed, 14 insertions(+), 13 deletions(-)

diff --git 
a/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/CacheGroupsMetricsRebalanceTest.java
 
b/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/CacheGroupsMetricsRebalanceTest.java
index ddffe356010..aaec37f1532 100644
--- 
a/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/CacheGroupsMetricsRebalanceTest.java
+++ 
b/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/CacheGroupsMetricsRebalanceTest.java
@@ -19,6 +19,7 @@ package org.apache.ignite.internal.processors.cache;
 
 import java.util.Arrays;
 import java.util.Collection;
+import java.util.Collections;
 import java.util.List;
 import java.util.Map;
 import java.util.Random;
@@ -38,7 +39,6 @@ import 
org.apache.ignite.cache.affinity.rendezvous.RendezvousAffinityFunction;
 import org.apache.ignite.cluster.ClusterNode;
 import org.apache.ignite.configuration.CacheConfiguration;
 import org.apache.ignite.configuration.IgniteConfiguration;
-import org.apache.ignite.events.Event;
 import org.apache.ignite.events.EventType;
 import org.apache.ignite.internal.IgniteEx;
 import org.apache.ignite.internal.IgniteInternalFuture;
@@ -51,7 +51,6 @@ import org.apache.ignite.internal.util.typedef.internal.A;
 import org.apache.ignite.internal.util.typedef.internal.CU;
 import org.apache.ignite.internal.util.typedef.internal.U;
 import org.apache.ignite.lang.IgniteBiPredicate;
-import org.apache.ignite.lang.IgnitePredicate;
 import org.apache.ignite.metric.MetricRegistry;
 import org.apache.ignite.plugin.extensions.communication.Message;
 import org.apache.ignite.spi.metric.IntMetric;
@@ -427,7 +426,19 @@ public class CacheGroupsMetricsRebalanceTest extends 
GridCommonAbstractTest {
             }
         });
 
-        IgniteEx ig = startGrid(4);
+        CountDownLatch latch = new CountDownLatch(1);
+
+        // Register the listener via node configuration so it is active before 
the node joins and starts rebalancing:
+        // otherwise EVT_CACHE_REBALANCE_STOPPED may fire before localListen() 
and the latch would never be released.
+        IgniteConfiguration cfg = 
optimize(getConfiguration(getTestIgniteInstanceName(4)));
+
+        cfg.setLocalEventListeners(Collections.singletonMap(evt -> {
+            latch.countDown();
+
+            return false;
+        }, new int[] {EventType.EVT_CACHE_REBALANCE_STOPPED}));
+
+        IgniteEx ig = startGrid(cfg);
 
         GridTestUtils.runAsync(new Runnable() {
             @Override public void run() {
@@ -439,16 +450,6 @@ public class CacheGroupsMetricsRebalanceTest extends 
GridCommonAbstractTest {
             }
         });
 
-        CountDownLatch latch = new CountDownLatch(1);
-
-        ig.events().localListen(new IgnitePredicate<Event>() {
-            @Override public boolean apply(Event evt) {
-                latch.countDown();
-
-                return false;
-            }
-        }, EventType.EVT_CACHE_REBALANCE_STOPPED);
-
         latch.await();
 
         GridTestUtils.waitForCondition(new GridAbsPredicate() {