diff --git a/core/src/main/scala/kafka/cluster/Partition.scala b/core/src/main/scala/kafka/cluster/Partition.scala
index a2c18c241d..7329eae2e6 100755
--- a/core/src/main/scala/kafka/cluster/Partition.scala
+++ b/core/src/main/scala/kafka/cluster/Partition.scala
@@ -572,6 +572,13 @@ class Partition(val topicPartition: TopicPartition,
   }
 
   // AutoMQ for Kafka inject start
+  private def checkClosed(): Unit = {
+    if (closed) {
+      throw new NotLeaderOrFollowerException("Leader %d for partition %s on broker %d is already closed"
+        .format(localBrokerId, topicPartition, localBrokerId))
+    }
+  }
+
   /**
    * Close this partition and trigger a re-election.
    * It may be messy here, but it is necessary to ensure the following steps:
@@ -1268,12 +1275,10 @@ class Partition(val topicPartition: TopicPartition,
 
   def appendRecordsToLeader(records: MemoryRecords, origin: AppendOrigin, requiredAcks: Int,
                             requestLocal: RequestLocal): LogAppendInfo = {
+    checkClosed()
     val (info, leaderHWIncremented) = inReadLock(leaderIsrUpdateLock) {
       // AutoMQ for Kafka inject start
-      if (closed) {
-        throw new NotLeaderOrFollowerException("Leader %d for partition %s on broker %d is already closed"
-          .format(localBrokerId, topicPartition, localBrokerId))
-      }
+      checkClosed()
       // AutoMQ for Kafka inject end
       leaderLogIfLocal match {
         case Some(leaderLog) =>
@@ -1372,7 +1377,9 @@ class Partition(val topicPartition: TopicPartition,
 
       logReadInfo
     } else {
+      checkClosed()
       inReadLock(leaderIsrUpdateLock) {
+        checkClosed()
         val localLog = localLogWithEpochOrThrow(
           fetchPartitionData.currentLeaderEpoch,
           fetchParams.fetchOnlyLeader