[kafka] branch trunk updated: KAFKA-7313; StopReplicaRequest should attempt to remove future replica for the partition only if future replica exists

[li...@apache.org]

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

lindong pushed a commit to branch trunk
in repository https://gitbox.apache.org/repos/asf/kafka.git


The following commit(s) were added to refs/heads/trunk by this push:
     new 6f83d05  KAFKA-7313; StopReplicaRequest should attempt to remove future replica for the partition only if future replica exists
6f83d05 is described below

commit 6f83d051315e120d148a53e3deba6541d3bb9857
Author: Dong Lin <li...@gmail.com>
AuthorDate: Tue Nov 6 11:40:39 2018 -0800

    KAFKA-7313; StopReplicaRequest should attempt to remove future replica for the partition only if future replica exists
    
    This patch fixes two issues:
    
    1) Currently if a broker received StopReplicaRequest with delete=true for the same offline replica, the first StopRelicaRequest will show KafkaStorageException and the second StopRelicaRequest will show ReplicaNotAvailableException. This is because the first StopRelicaRequest will remove the mapping (tp -> ReplicaManager.OfflinePartition) from ReplicaManager.allPartitions before returning KafkaStorageException, thus the second StopRelicaRequest will not find this partition as offline.
    
    This result appears to be inconsistent. And since the replica is already offline and broker will not be able to delete file for this replica, the StopReplicaRequest should fail without making any change and broker should still remember that this replica is offline.
    
    2) Currently if broker receives StopReplicaRequest with delete=true, the broker will attempt to remove future replica for the partition, which will cause KafkaStorageException in the StopReplicaResponse if this replica does not have future replica. It is problematic to always return KafkaStorageException in the response if future replica does not exist.
    
    Author: Dong Lin <li...@gmail.com>
    
    Reviewers: Jun Rao <ju...@gmail.com>
    
    Closes #5533 from lindong28/KAFKA-7313
---
 core/src/main/scala/kafka/cluster/Partition.scala  |  3 +-
 core/src/main/scala/kafka/log/LogManager.scala     |  2 +-
 .../main/scala/kafka/server/ReplicaManager.scala   | 11 +++--
 .../unit/kafka/server/StopReplicaRequestTest.scala | 57 ++++++++++++++++++++++
 4 files changed, 68 insertions(+), 5 deletions(-)

diff --git a/core/src/main/scala/kafka/cluster/Partition.scala b/core/src/main/scala/kafka/cluster/Partition.scala
index 819da2c..745c89a 100755
--- a/core/src/main/scala/kafka/cluster/Partition.scala
+++ b/core/src/main/scala/kafka/cluster/Partition.scala
@@ -352,7 +352,8 @@ class Partition(val topicPartition: TopicPartition,
       leaderEpochStartOffsetOpt = None
       removePartitionMetrics()
       logManager.asyncDelete(topicPartition)
-      logManager.asyncDelete(topicPartition, isFuture = true)
+      if (logManager.getLog(topicPartition, isFuture = true).isDefined)
+        logManager.asyncDelete(topicPartition, isFuture = true)
     }
   }
 
diff --git a/core/src/main/scala/kafka/log/LogManager.scala b/core/src/main/scala/kafka/log/LogManager.scala
index 39029b0..26bfbe9 100755
--- a/core/src/main/scala/kafka/log/LogManager.scala
+++ b/core/src/main/scala/kafka/log/LogManager.scala
@@ -845,7 +845,7 @@ class LogManager(logDirs: Seq[File],
       addLogToBeDeleted(removedLog)
       info(s"Log for partition ${removedLog.topicPartition} is renamed to ${removedLog.dir.getAbsolutePath} and is scheduled for deletion")
     } else if (offlineLogDirs.nonEmpty) {
-      throw new KafkaStorageException("Failed to delete log for " + topicPartition + " because it may be in one of the offline directories " + offlineLogDirs.mkString(","))
+      throw new KafkaStorageException(s"Failed to delete log for ${if (isFuture) "future" else ""} $topicPartition because it may be in one of the offline directories ${offlineLogDirs.mkString(",")}")
     }
     removedLog
   }
diff --git a/core/src/main/scala/kafka/server/ReplicaManager.scala b/core/src/main/scala/kafka/server/ReplicaManager.scala
index 1146bef..443a5cf 100644
--- a/core/src/main/scala/kafka/server/ReplicaManager.scala
+++ b/core/src/main/scala/kafka/server/ReplicaManager.scala
@@ -338,8 +338,10 @@ class ReplicaManager(val config: KafkaConfig,
 
     if (deletePartition) {
       val removedPartition = allPartitions.remove(topicPartition)
-      if (removedPartition eq ReplicaManager.OfflinePartition)
+      if (removedPartition eq ReplicaManager.OfflinePartition) {
+        allPartitions.put(topicPartition, ReplicaManager.OfflinePartition)
         throw new KafkaStorageException(s"Partition $topicPartition is on an offline disk")
+      }
 
       if (removedPartition != null) {
         val topicHasPartitions = allPartitions.values.exists(partition => topicPartition.topic == partition.topic)
@@ -1402,7 +1404,8 @@ class ReplicaManager(val config: KafkaConfig,
   }
 
   // logDir should be an absolute path
-  def handleLogDirFailure(dir: String) {
+  // sendZkNotification is needed for unit test
+  def handleLogDirFailure(dir: String, sendZkNotification: Boolean = true) {
     if (!logManager.isLogDirOnline(dir))
       return
     info(s"Stopping serving replicas in dir $dir")
@@ -1438,7 +1441,9 @@ class ReplicaManager(val config: KafkaConfig,
            s"for partitions ${partitionsWithOfflineFutureReplica.mkString(",")} because they are in the failed log directory $dir.")
     }
     logManager.handleLogDirFailure(dir)
-    zkClient.propagateLogDirEvent(localBrokerId)
+
+    if (sendZkNotification)
+      zkClient.propagateLogDirEvent(localBrokerId)
     info(s"Stopped serving replicas in dir $dir")
   }
 
diff --git a/core/src/test/scala/unit/kafka/server/StopReplicaRequestTest.scala b/core/src/test/scala/unit/kafka/server/StopReplicaRequestTest.scala
new file mode 100644
index 0000000..5df61eb
--- /dev/null
+++ b/core/src/test/scala/unit/kafka/server/StopReplicaRequestTest.scala
@@ -0,0 +1,57 @@
+/**
+  * Licensed to the Apache Software Foundation (ASF) under one or more
+  * contributor license agreements.  See the NOTICE file distributed with
+  * this work for additional information regarding copyright ownership.
+  * The ASF licenses this file to You under the Apache License, Version 2.0
+  * (the "License"); you may not use this file except in compliance with
+  * the License.  You may obtain a copy of the License at
+  *
+  *    http://www.apache.org/licenses/LICENSE-2.0
+  *
+  * Unless required by applicable law or agreed to in writing, software
+  * distributed under the License is distributed on an "AS IS" BASIS,
+  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+  * See the License for the specific language governing permissions and
+  * limitations under the License.
+  */
+
+package kafka.server
+
+import kafka.utils._
+import org.apache.kafka.common.TopicPartition
+import org.apache.kafka.common.protocol.{ApiKeys, Errors}
+import org.apache.kafka.common.requests._
+import org.junit.Assert._
+import org.junit.Test
+import collection.JavaConverters._
+
+
+class StopReplicaRequestTest extends BaseRequestTest {
+  override val logDirCount = 2
+  override val numBrokers: Int = 1
+
+  val topic = "topic"
+  val partitionNum = 2
+  val tp0 = new TopicPartition(topic, 0)
+  val tp1 = new TopicPartition(topic, 1)
+
+  @Test
+  def testStopReplicaRequest(): Unit = {
+    createTopic(topic, partitionNum, 1)
+    TestUtils.generateAndProduceMessages(servers, topic, 10)
+
+    val server = servers.head
+    val offlineDir = server.logManager.getLog(tp1).get.dir.getParent
+    server.replicaManager.handleLogDirFailure(offlineDir, sendZkNotification = false)
+
+    for (i <- 1 to 2) {
+      val request1 = new StopReplicaRequest.Builder(
+        server.config.brokerId, server.replicaManager.controllerEpoch, true, Set(tp0, tp1).asJava).build()
+      val response1 = connectAndSend(request1, ApiKeys.STOP_REPLICA, controllerSocketServer)
+      val partitionErrors1 = StopReplicaResponse.parse(response1, request1.version).responses()
+      assertEquals(Errors.NONE, partitionErrors1.get(tp0))
+      assertEquals(Errors.KAFKA_STORAGE_ERROR, partitionErrors1.get(tp1))
+    }
+  }
+
+}