You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pulsar.apache.org by pe...@apache.org on 2022/06/28 15:13:19 UTC

[pulsar] 14/29: [fix][txn] Fix NPE when ack message with transaction at cnx = null (#16142)

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

penghui pushed a commit to branch branch-2.10
in repository https://gitbox.apache.org/repos/asf/pulsar.git

commit c5a6a0b78b3538098eb04466c8242128144011a9
Author: Xiangying Meng <55...@users.noreply.github.com>
AuthorDate: Tue Jun 21 15:33:23 2022 +0800

    [fix][txn] Fix NPE when ack message with transaction at cnx = null  (#16142)
    
    Fix https://github.com/apache/pulsar/issues/16124
    ## Motivation
    When a channel is inactive, connectHandler will set the cnx = null and reconnect.
    At this time, consumers use transaction to ack messages will report NPE.
    ## Modification
    Return exception when cnx = null.
    
    **Why not use a queue to store operations?**
    1. If we use a queue to store op, we need to take care of the timeout of the op. And the lock is required.
    2. If the connection time is long or there is a BUG client that has not been connected, the client will crash.
    
    (cherry picked from commit 53cc84a580dd747685905e1d11b8e19c0e59a614)
---
 .../pulsar/broker/transaction/TransactionTest.java | 41 ++++++++++++++++++++++
 .../apache/pulsar/client/impl/ConsumerImpl.java    |  9 ++++-
 2 files changed, 49 insertions(+), 1 deletion(-)

diff --git a/pulsar-broker/src/test/java/org/apache/pulsar/broker/transaction/TransactionTest.java b/pulsar-broker/src/test/java/org/apache/pulsar/broker/transaction/TransactionTest.java
index 3f6ea313652..68e73d895e3 100644
--- a/pulsar-broker/src/test/java/org/apache/pulsar/broker/transaction/TransactionTest.java
+++ b/pulsar-broker/src/test/java/org/apache/pulsar/broker/transaction/TransactionTest.java
@@ -51,6 +51,7 @@ import java.util.concurrent.TimeoutException;
 import java.util.concurrent.atomic.AtomicLong;
 import lombok.Cleanup;
 import lombok.extern.slf4j.Slf4j;
+import org.apache.bookkeeper.common.util.Bytes;
 import org.apache.bookkeeper.mledger.AsyncCallbacks;
 import org.apache.bookkeeper.mledger.ManagedCursor;
 import org.apache.bookkeeper.mledger.ManagedLedgerException;
@@ -87,6 +88,7 @@ import org.apache.pulsar.client.api.Schema;
 import org.apache.pulsar.client.api.SubscriptionType;
 import org.apache.pulsar.client.api.transaction.Transaction;
 import org.apache.pulsar.client.api.transaction.TxnID;
+import org.apache.pulsar.client.impl.ClientCnx;
 import org.apache.pulsar.client.impl.MessageIdImpl;
 import org.apache.pulsar.common.api.proto.CommandSubscribe;
 import org.apache.pulsar.client.impl.transaction.TransactionImpl;
@@ -109,6 +111,7 @@ import org.apache.pulsar.transaction.coordinator.impl.MLTransactionLogImpl;
 import org.apache.pulsar.transaction.coordinator.impl.MLTransactionSequenceIdGenerator;
 import org.apache.pulsar.transaction.coordinator.impl.MLTransactionMetadataStore;
 import org.awaitility.Awaitility;
+import org.powermock.reflect.Whitebox;
 import org.testng.Assert;
 import org.testng.annotations.AfterMethod;
 import org.testng.annotations.BeforeMethod;
@@ -992,6 +995,44 @@ public class TransactionTest extends TransactionTestBase {
         transaction.commit().get();
     }
 
+    @Test
+    public void testGetConnectExceptionForAckMsgWhenCnxIsNull() throws Exception {
+        String topic = NAMESPACE1 + "/testGetConnectExceptionForAckMsgWhenCnxIsNull";
+        @Cleanup
+        Producer<byte[]> producer = pulsarClient
+                .newProducer(Schema.BYTES)
+                .topic(topic)
+                .sendTimeout(0, TimeUnit.SECONDS)
+                .create();
+
+        @Cleanup
+        Consumer<byte[]> consumer = pulsarClient
+                .newConsumer()
+                .topic(topic)
+                .subscriptionName("sub")
+                .subscribe();
+
+        for (int i = 0; i < 10; i++) {
+            producer.newMessage().value(Bytes.toBytes(i)).send();
+        }
+        ClientCnx cnx = Whitebox.invokeMethod(consumer, "cnx");
+        Whitebox.invokeMethod(consumer, "connectionClosed", cnx);
+
+        Message<byte[]> message = consumer.receive();
+        Transaction transaction = pulsarClient
+                .newTransaction()
+                .withTransactionTimeout(5, TimeUnit.SECONDS)
+                .build().get();
+
+        try {
+            consumer.acknowledgeAsync(message.getMessageId(), transaction).get();
+            fail();
+        } catch (ExecutionException e) {
+            Assert.assertTrue(e.getCause() instanceof PulsarClientException.ConnectException);
+        }
+    }
+
+
     @Test
     public void testPendingAckBatchMessageCommit() throws Exception {
         String topic = NAMESPACE1 + "/testPendingAckBatchMessageCommit";
diff --git a/pulsar-client/src/main/java/org/apache/pulsar/client/impl/ConsumerImpl.java b/pulsar-client/src/main/java/org/apache/pulsar/client/impl/ConsumerImpl.java
index b4ee5a2e784..9422877ef63 100644
--- a/pulsar-client/src/main/java/org/apache/pulsar/client/impl/ConsumerImpl.java
+++ b/pulsar-client/src/main/java/org/apache/pulsar/client/impl/ConsumerImpl.java
@@ -2644,7 +2644,14 @@ public class ConsumerImpl<T> extends ConsumerBase<T> implements ConnectionHandle
         } else {
             unAckedMessageTracker.remove(messageId);
         }
-        return cnx().newAckForReceipt(cmd, requestId);
+        ClientCnx cnx = cnx();
+        if (cnx == null) {
+            return FutureUtil.failedFuture(new PulsarClientException
+                    .ConnectException("Failed to ack message [" + messageId + "] "
+                    + "for transaction [" + txnID + "] due to consumer connect fail, consumer state: " + getState()));
+        } else {
+            return cnx.newAckForReceipt(cmd, requestId);
+        }
     }
 
     public Map<MessageIdImpl, List<MessageImpl<T>>> getPossibleSendToDeadLetterTopicMessages() {