HDFS-14397. Backport HADOOP-15684 to branch-2. Contributed by Chao Sun.

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/ha/EditLogTailer.java

@@ -57,8 +57,6 @@ import org.apache.hadoop.hdfs.server.namenode.FSNamesystem;
 import org.apache.hadoop.hdfs.server.namenode.NameNode;
 import org.apache.hadoop.hdfs.server.protocol.NamenodeProtocol;
 import org.apache.hadoop.ipc.RPC;
-import org.apache.hadoop.ipc.RemoteException;
-import org.apache.hadoop.ipc.StandbyException;
 import org.apache.hadoop.security.SecurityUtil;
 
 import static org.apache.hadoop.util.Time.monotonicNow;
@@ -344,15 +342,6 @@ public class EditLogTailer {
       future.get(rollEditsTimeoutMs, TimeUnit.MILLISECONDS);
       lastRollTriggerTxId = lastLoadedTxnId;
     } catch (ExecutionException e) {
-      Throwable cause = e.getCause();
-      if (cause instanceof RemoteException) {
-        IOException ioe = ((RemoteException) cause).unwrapRemoteException();
-        if (ioe instanceof StandbyException) {
-          LOG.info("Skipping log roll. Remote node is not in Active state: " +
-              ioe.getMessage().split("\n")[0]);
-          return;
-        }
-      }
       LOG.warn("Unable to trigger a roll of the active NN", e);
     } catch (TimeoutException e) {
       if (future != null) {
@@ -455,7 +444,8 @@ public class EditLogTailer {
    * This mechanism is <b>very bad</b> for cases where we care about being <i>fast</i>; it just
    * blindly goes and tries namenodes.
    */
-  private abstract class MultipleNameNodeProxy<T> implements Callable<T> {
+  @VisibleForTesting
+  abstract class MultipleNameNodeProxy<T> implements Callable<T> {
 
     /**
      * Do the actual work to the remote namenode via the {@link #cachedActiveProxy}.
@@ -471,19 +461,13 @@ public class EditLogTailer {
         try {
           T ret = doWork();
           return ret;
-        } catch (RemoteException e) {
-          Throwable cause = e.unwrapRemoteException(StandbyException.class);
-          // if its not a standby exception, then we need to re-throw it, something bad has happened
-          if (cause == e) {
-            throw e;
-          } else {
-            // it is a standby exception, so we try the other NN
-            LOG.warn("Failed to reach remote node: " + currentNN
-                + ", retrying with remaining remote NNs");
-            cachedActiveProxy = null;
-            // this NN isn't responding to requests, try the next one
-            nnLoopCount++;
-          }
+        } catch (IOException e) {
+          LOG.warn("Exception from remote name node " + currentNN
+              + ", try next.", e);
+
+          // Try next name node if exception happens.
+          cachedActiveProxy = null;
+          nnLoopCount++;
         }
       }
       throw new IOException("Cannot find any valid remote NN to service request!");

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/MiniDFSNNTopology.java

@@ -72,6 +72,23 @@ public class MiniDFSNNTopology {
     return topology;
   }
 
+  /**
+   * Set up an HA topology with a single HA nameservice.
+   * @param nnCount of namenodes to use with the nameservice
+   * @param basePort for IPC and Http ports of namenodes.
+   */
+  public static MiniDFSNNTopology simpleHATopology(int nnCount, int basePort) {
+    MiniDFSNNTopology.NSConf ns = new MiniDFSNNTopology.NSConf("minidfs-ns");
+    for (int i = 0; i < nnCount; i++) {
+      ns.addNN(new MiniDFSNNTopology.NNConf("nn" + i)
+          .setIpcPort(basePort++)
+          .setHttpPort(basePort++));
+    }
+    MiniDFSNNTopology topology = new MiniDFSNNTopology()
+        .addNameservice(ns);
+    return topology;
+  }
+
   /**
    * Set up federated cluster with the given number of nameservices, each
    * of which has only a single NameNode.

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/namenode/ha/TestEditLogTailer.java

@@ -28,6 +28,7 @@ import java.net.BindException;
 import java.net.URI;
 import java.util.ArrayList;
 import java.util.Collection;
+import java.util.Random;
 import java.util.concurrent.Callable;
 import java.util.concurrent.atomic.AtomicInteger;
 import java.util.concurrent.TimeoutException;
@@ -46,7 +47,6 @@ import org.apache.hadoop.hdfs.server.namenode.FSImage;
 import org.apache.hadoop.hdfs.server.namenode.NNStorage;
 import org.apache.hadoop.hdfs.server.namenode.NameNode;
 import org.apache.hadoop.hdfs.server.namenode.NameNodeAdapter;
-import org.apache.hadoop.net.ServerSocketUtil;
 import org.apache.hadoop.test.GenericTestUtils;
 import org.apache.log4j.Level;
 import org.junit.Test;
@@ -168,21 +168,7 @@ public class TestEditLogTailer {
     MiniDFSCluster cluster = null;
     for (int i = 0; i < 5; i++) {
       try {
-        // Have to specify IPC ports so the NNs can talk to each other.
-        int[] ports = ServerSocketUtil.getPorts(3);
-        MiniDFSNNTopology topology = new MiniDFSNNTopology()
-            .addNameservice(new MiniDFSNNTopology.NSConf("ns1")
-                .addNN(new MiniDFSNNTopology.NNConf("nn1")
-                    .setIpcPort(ports[0]))
-                .addNN(new MiniDFSNNTopology.NNConf("nn2")
-                    .setIpcPort(ports[1]))
-                .addNN(new MiniDFSNNTopology.NNConf("nn3")
-                    .setIpcPort(ports[2])));
-
-        cluster = new MiniDFSCluster.Builder(conf)
-          .nnTopology(topology)
-          .numDataNodes(0)
-          .build();
+        cluster = createMiniDFSCluster(conf, 3);
         break;
       } catch (BindException e) {
         // retry if race on ports given by ServerSocketUtil#getPorts
@@ -213,21 +199,9 @@ public class TestEditLogTailer {
     conf.setInt(DFSConfigKeys.DFS_HA_TAILEDITS_PERIOD_KEY, 1);
     conf.setInt(DFSConfigKeys.DFS_HA_TAILEDITS_ALL_NAMESNODES_RETRY_KEY, 100);
 
-    // Have to specify IPC ports so the NNs can talk to each other.
-    MiniDFSNNTopology topology = new MiniDFSNNTopology()
-        .addNameservice(new MiniDFSNNTopology.NSConf("ns1")
-            .addNN(new MiniDFSNNTopology.NNConf("nn1")
-                .setIpcPort(ServerSocketUtil.getPort(0, 100)))
-            .addNN(new MiniDFSNNTopology.NNConf("nn2")
-                .setIpcPort(ServerSocketUtil.getPort(0, 100)))
-            .addNN(new MiniDFSNNTopology.NNConf("nn3")
-                .setIpcPort(ServerSocketUtil.getPort(0, 100))));
-
-    MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf)
-        .nnTopology(topology)
-        .numDataNodes(0)
-        .build();
+    MiniDFSCluster cluster = null;
     try {
+      cluster = createMiniDFSCluster(conf, 3);
       cluster.transitionToStandby(0);
       cluster.transitionToStandby(1);
       cluster.transitionToStandby(2);
@@ -240,7 +214,9 @@ public class TestEditLogTailer {
       cluster.transitionToActive(0);
       waitForLogRollInSharedDir(cluster, 3);
     } finally {
-      cluster.shutdown();
+      if (cluster != null) {
+        cluster.shutdown();
+      }
     }
   }
   
@@ -306,4 +282,64 @@ public class TestEditLogTailer {
       cluster.shutdown();
     }
   }
+
+  @Test
+  public void testRollEditLogIOExceptionForRemoteNN() throws IOException {
+    Configuration conf = getConf();
+
+    // Roll every 1s
+    conf.setInt(DFSConfigKeys.DFS_HA_LOGROLL_PERIOD_KEY, 1);
+    conf.setInt(DFSConfigKeys.DFS_HA_TAILEDITS_PERIOD_KEY, 1);
+
+    MiniDFSCluster cluster = null;
+    try {
+      cluster = createMiniDFSCluster(conf, 3);
+      cluster.transitionToActive(0);
+      EditLogTailer tailer = Mockito.spy(
+          cluster.getNamesystem(1).getEditLogTailer());
+
+      final AtomicInteger invokedTimes = new AtomicInteger(0);
+
+      // It should go on to next name node when IOException happens.
+      when(tailer.getNameNodeProxy()).thenReturn(
+          tailer.new MultipleNameNodeProxy<Void>() {
+            @Override
+            protected Void doWork() throws IOException {
+              invokedTimes.getAndIncrement();
+              throw new IOException("It is an IO Exception.");
+            }
+          }
+      );
+
+      tailer.triggerActiveLogRoll();
+
+      // MultipleNameNodeProxy uses Round-robin to look for active NN
+      // to do RollEditLog. If doWork() fails, then IOException throws,
+      // it continues to try next NN. triggerActiveLogRoll finishes
+      // either due to success, or using up retries.
+      // In this test case, there are 2 remote name nodes, default retry is 3.
+      // For test purpose, doWork() always returns IOException,
+      // so the total invoked times will be default retry 3 * remote NNs 2 = 6
+      assertEquals(6, invokedTimes.get());
+    } finally {
+      if (cluster != null) {
+        cluster.shutdown();
+      }
+    }
+  }
+
+  private static MiniDFSCluster createMiniDFSCluster(Configuration conf,
+      int nnCount) throws IOException {
+    int basePort = 10060 + new Random().nextInt(100) * 2;
+
+    // By passing in basePort, name node will have IPC port set,
+    // which is needed for enabling roll log.
+    MiniDFSNNTopology topology =
+            MiniDFSNNTopology.simpleHATopology(nnCount, basePort);
+    MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf)
+        .nnTopology(topology)
+        .numDataNodes(0)
+        .build();
+    return cluster;
+  }
 }