Use unwrapped cause to determine if node is closing (#39723)

We need to unwrap and use the actual cause when determining if the node with primary shard is shutting down because TransportService will throw a TransportException wrapped in a SendRequestTransportException. Relates #39584
elastic · Mar 6, 2019 · 71c64bf · 71c64bf
1 parent 41706c3
commit 71c64bf
Show file tree

Hide file tree

Showing 3 changed files with 14 additions and 6 deletions.
diff --git a/server/src/main/java/org/elasticsearch/action/support/replication/ReplicationOperation.java b/server/src/main/java/org/elasticsearch/action/support/replication/ReplicationOperation.java
@@ -199,8 +199,9 @@ public void onFailure(Exception replicaException) {
     }
 
     private void onNoLongerPrimary(Exception failure) {
-        final boolean nodeIsClosing = failure instanceof NodeClosedException ||
-            (failure instanceof TransportException && "TransportService is closed stopped can't send request".equals(failure.getMessage()));
+        final Throwable cause = ExceptionsHelper.unwrapCause(failure);
+        final boolean nodeIsClosing = cause instanceof NodeClosedException
+            || (cause instanceof TransportException && "TransportService is closed stopped can't send request".equals(cause.getMessage()));
         final String message;
         if (nodeIsClosing) {
             message = String.format(Locale.ROOT,

diff --git a/...src/test/java/org/elasticsearch/action/support/replication/ReplicationOperationTests.java b/...src/test/java/org/elasticsearch/action/support/replication/ReplicationOperationTests.java
@@ -43,6 +43,7 @@
 import org.elasticsearch.index.shard.ShardId;
 import org.elasticsearch.node.NodeClosedException;
 import org.elasticsearch.test.ESTestCase;
+import org.elasticsearch.transport.SendRequestTransportException;
 import org.elasticsearch.transport.TransportException;
 
 import java.util.ArrayList;
@@ -203,7 +204,9 @@ public void testNoLongerPrimary() throws Exception {
         if (randomBoolean()) {
             shardActionFailure = new NodeClosedException(new DiscoveryNode("foo", buildNewFakeTransportAddress(), Version.CURRENT));
         } else if (randomBoolean()) {
-            shardActionFailure = new TransportException("TransportService is closed stopped can't send request");
+            shardActionFailure = new SendRequestTransportException(
+                new DiscoveryNode("foo", buildNewFakeTransportAddress(), Version.CURRENT), "internal:cluster/shard/failure",
+                new TransportException("TransportService is closed stopped can't send request"));
         } else {
             shardActionFailure = new ShardStateAction.NoLongerPrimaryShardException(failedReplica.shardId(), "the king is dead");
         }

diff --git a/server/src/test/java/org/elasticsearch/discovery/ClusterDisruptionIT.java b/server/src/test/java/org/elasticsearch/discovery/ClusterDisruptionIT.java
@@ -70,8 +70,10 @@
 import static org.elasticsearch.common.xcontent.XContentFactory.jsonBuilder;
 import static org.elasticsearch.test.hamcrest.ElasticsearchAssertions.assertAcked;
 import static org.hamcrest.Matchers.equalTo;
+import static org.hamcrest.Matchers.everyItem;
 import static org.hamcrest.Matchers.greaterThanOrEqualTo;
 import static org.hamcrest.Matchers.is;
+import static org.hamcrest.Matchers.isIn;
 import static org.hamcrest.Matchers.isOneOf;
 import static org.hamcrest.Matchers.not;
 
@@ -453,7 +455,7 @@ public void testRestartNodeWhileIndexing() throws Exception {
                     String id = Integer.toString(docID.incrementAndGet());
                     try {
                         IndexResponse response = client().prepareIndex(index, "_doc", id).setSource("{}", XContentType.JSON).get();
-                        assertThat(response.getResult(), isOneOf(CREATED, UPDATED));
+                        assertThat(response.getResult(), isOneOf(DocWriteResponse.Result.CREATED, DocWriteResponse.Result.UPDATED));
                         logger.info("--> index id={} seq_no={}", response.getId(), response.getSeqNo());
                         ackedDocs.add(response.getId());
                     } catch (ElasticsearchException ignore) {
@@ -476,8 +478,10 @@ public void testRestartNodeWhileIndexing() throws Exception {
         for (ShardRouting shardRouting : clusterState.routingTable().allShards(index)) {
             String nodeName = clusterState.nodes().get(shardRouting.currentNodeId()).getName();
             IndicesService indicesService = internalCluster().getInstance(IndicesService.class, nodeName);
-            IndexShard indexShard = indicesService.getShardOrNull(shardRouting.shardId());
-            assertThat(IndexShardTestCase.getShardDocUIDs(indexShard), equalTo(ackedDocs));
+            IndexShard shard = indicesService.getShardOrNull(shardRouting.shardId());
+            Set<String> docs = IndexShardTestCase.getShardDocUIDs(shard);
+            assertThat("shard [" + shard.routingEntry() + "] docIds [" + docs + "] vs " + " acked docIds [" + ackedDocs + "]",
+                ackedDocs, everyItem(isIn(docs)));
         }
     }
 }