HDFS-8995. Flaw in registration bookeeping can make DN die on reconne…

…ct. (Kihwal Lee via yliu)
apache · Sep 2, 2015 · 5652131 · 5652131
1 parent 4620767
commit 5652131
Show file tree

Hide file tree

Showing 4 changed files with 9 additions and 9 deletions.
diff --git a/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt b/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt
@@ -1302,6 +1302,9 @@ Release 2.7.2 - UNRELEASED
 
     HDFS-8891. HDFS concat should keep srcs order. (Yong Zhang via jing9)
 
+    HDFS-8995. Flaw in registration bookeeping can make DN die on reconnect.
+    (Kihwal Lee via yliu)
+
 Release 2.7.1 - 2015-07-06
 
   INCOMPATIBLE CHANGES

diff --git a/...ject/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BPOfferService.java b/...ject/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BPOfferService.java
@@ -351,9 +351,8 @@ void registrationSucceeded(BPServiceActor bpServiceActor,
             reg.getStorageInfo().getNamespaceID(), "namespace ID");
         checkNSEquality(bpRegistration.getStorageInfo().getClusterID(),
             reg.getStorageInfo().getClusterID(), "cluster ID");
-      } else {
-        bpRegistration = reg;
       }
+      bpRegistration = reg;
 
       dn.bpRegistrationSucceeded(bpRegistration, getBlockPoolId());
       // Add the initial block token secret keys to the DN's secret manager.

diff --git a/...ject/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BPServiceActor.java b/...ject/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BPServiceActor.java
@@ -767,15 +767,16 @@ private void offerService() throws Exception {
   void register(NamespaceInfo nsInfo) throws IOException {
     // The handshake() phase loaded the block pool storage
     // off disk - so update the bpRegistration object from that info
-    bpRegistration = bpos.createRegistration();
+    DatanodeRegistration newBpRegistration = bpos.createRegistration();
 
     LOG.info(this + " beginning handshake with NN");
 
     while (shouldRun()) {
       try {
         // Use returned registration from namenode with updated fields
-        bpRegistration = bpNamenode.registerDatanode(bpRegistration);
-        bpRegistration.setNamespaceInfo(nsInfo);
+        newBpRegistration = bpNamenode.registerDatanode(newBpRegistration);
+        newBpRegistration.setNamespaceInfo(nsInfo);
+        bpRegistration = newBpRegistration;
         break;
       } catch(EOFException e) {  // namenode might have just restarted
         LOG.info("Problem connecting to server: " + nnAddr + " :"

diff --git a/...fs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DataNode.java b/...fs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DataNode.java
@@ -1261,10 +1261,7 @@ DatanodeRegistration createBPRegistration(NamespaceInfo nsInfo) {
    */
   synchronized void bpRegistrationSucceeded(DatanodeRegistration bpRegistration,
       String blockPoolId) throws IOException {
-    // Set the ID if we haven't already
-    if (null == id) {
-      id = bpRegistration;
-    }
+    id = bpRegistration;
 
     if(!storage.getDatanodeUuid().equals(bpRegistration.getDatanodeUuid())) {
       throw new IOException("Inconsistent Datanode IDs. Name-node returned "