HDFS-8929. Add a metric to expose the timestamp of the last journal (Contributed by surendra singh lilhore)

hadoop-common-project/hadoop-common/src/site/markdown/Metrics.md

@@ -275,6 +275,7 @@ The server-side metrics for a journal from the JournalNode's perspective. Each m
 | `CurrentLagTxns` | The number of transactions that this JournalNode is lagging |
 | `LastWrittenTxId` | The highest transaction id stored on this JournalNode |
 | `LastPromisedEpoch` | The last epoch number which this node has promised not to accept any lower epoch, or 0 if no promises have been made |
+| `LastJournalTimestamp` | The timestamp of last successfully written transaction |
 
 datanode
 --------

hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt

@@ -921,6 +921,9 @@ Release 2.8.0 - UNRELEASED
 
     HDFS-8862. BlockManager#excessReplicateMap should use a HashMap. (yliu)
 
+    HDFS-8929. Add a metric to expose the timestamp of the last journal
+    (surendra singh lilhore via vinayakumarb)
+
   BUG FIXES
 
     HDFS-7501. TransactionsSinceLastCheckpoint can be negative on SBNs.

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/qjournal/server/Journal.java

@@ -63,6 +63,7 @@
 import org.apache.hadoop.security.SecurityUtil;
 import org.apache.hadoop.security.UserGroupInformation;
 import org.apache.hadoop.util.StopWatch;
+import org.apache.hadoop.util.Time;
 
 import com.google.common.annotations.VisibleForTesting;
 import com.google.common.base.Charsets;
@@ -132,6 +133,8 @@ public class Journal implements Closeable {
 
   private final JournalMetrics metrics;
 
+  private long lastJournalTimestamp = 0;
+
   /**
    * Time threshold for sync calls, beyond which a warning should be logged to the console.
    */
@@ -254,6 +257,10 @@ synchronized long getCommittedTxnIdForTests() throws IOException {
     return committedTxnId.get();
   }
 
+  synchronized long getLastJournalTimestamp() {
+    return lastJournalTimestamp;
+  }
+
   synchronized long getCurrentLagTxns() throws IOException {
     long committed = committedTxnId.get();
     if (committed == 0) {
@@ -411,6 +418,7 @@ synchronized void journal(RequestInfo reqInfo,
     
     updateHighestWrittenTxId(lastTxnId);
     nextTxId = lastTxnId + 1;
+    lastJournalTimestamp = Time.now();
   }
 
   public void heartbeat(RequestInfo reqInfo) throws IOException {

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/qjournal/server/JournalMetrics.java

@@ -110,6 +110,11 @@ public long getCurrentLagTxns() {
     }
   }
 
+  @Metric("The timestamp of last successfully written transaction")
+  public long getLastJournalTimestamp() {
+    return journal.getLastJournalTimestamp();
+  }
+
   void addSync(long us) {
     for (MutableQuantiles q : syncsQuantiles) {
       q.add(us);

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/qjournal/server/TestJournalNode.java

@@ -106,7 +106,9 @@ public void testJournal() throws Exception {
     MetricsAsserts.assertCounter("BatchesWritten", 0L, metrics);
     MetricsAsserts.assertCounter("BatchesWrittenWhileLagging", 0L, metrics);
     MetricsAsserts.assertGauge("CurrentLagTxns", 0L, metrics);
+    MetricsAsserts.assertGauge("LastJournalTimestamp", 0L, metrics);
 
+    long beginTimestamp = System.currentTimeMillis();
     IPCLoggerChannel ch = new IPCLoggerChannel(
         conf, FAKE_NSINFO, journalId, jn.getBoundIpcAddress());
     ch.newEpoch(1).get();
@@ -119,6 +121,10 @@ public void testJournal() throws Exception {
     MetricsAsserts.assertCounter("BatchesWritten", 1L, metrics);
     MetricsAsserts.assertCounter("BatchesWrittenWhileLagging", 0L, metrics);
     MetricsAsserts.assertGauge("CurrentLagTxns", 0L, metrics);
+    long lastJournalTimestamp = MetricsAsserts.getLongGauge(
+        "LastJournalTimestamp", metrics);
+    assertTrue(lastJournalTimestamp > beginTimestamp);
+    beginTimestamp = lastJournalTimestamp;
 
     ch.setCommittedTxId(100L);
     ch.sendEdits(1L, 2, 1, "goodbye".getBytes(Charsets.UTF_8)).get();
@@ -128,6 +134,9 @@ public void testJournal() throws Exception {
     MetricsAsserts.assertCounter("BatchesWritten", 2L, metrics);
     MetricsAsserts.assertCounter("BatchesWrittenWhileLagging", 1L, metrics);
     MetricsAsserts.assertGauge("CurrentLagTxns", 98L, metrics);
+    lastJournalTimestamp = MetricsAsserts.getLongGauge(
+        "LastJournalTimestamp", metrics);
+    assertTrue(lastJournalTimestamp > beginTimestamp);
 
   }