svn commit: r1333200 - in /hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase: master/SplitLogManager.java zookeeper/ZKSplitLog.java

[mb...@apache.org]

Author: mbautin
Date: Wed May  2 20:53:40 2012
New Revision: 1333200

URL: http://svn.apache.org/viewvc?rev=1333200&view=rev
Log:
[HBASE-5860][jira] follow up diff

Author: pkhemani

Summary:
implement Nicolas's feedback

1 do not retry RESCAN creation
2 do not create RESCAN when other RESCAN create pending

Test Plan: -

Reviewers: nspiegelberg, kannan, aaiyer

Reviewed By: aaiyer

CC: hbase-eng@

Differential Revision: https://phabricator.fb.com/D461234

Modified:
    hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/master/SplitLogManager.java
    hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKSplitLog.java

Modified: hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/master/SplitLogManager.java
URL: http://svn.apache.org/viewvc/hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/master/SplitLogManager.java?rev=1333200&r1=1333199&r2=1333200&view=diff
==============================================================================
--- hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/master/SplitLogManager.java (original)
+++ hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/master/SplitLogManager.java Wed May  2 20:53:40 2012
@@ -557,7 +557,7 @@ public class SplitLogManager implements 
       task.unforcedResubmits++;
     }
     task.setUnassigned();
-    createRescanNode(Long.MAX_VALUE);
+    createRescanNode();
     tot_mgr_resubmit.incrementAndGet();
     return true;
   }
@@ -610,7 +610,7 @@ public class SplitLogManager implements 
    * signal the workers that a task was resubmitted by creating the
    * RESCAN node.
    */
-  private void createRescanNode(long retries) {
+  private void createRescanNode() {
     // The RESCAN node will be deleted almost immediately by the
     // SplitLogManager as soon as it is created because it is being
     // created in the DONE state. This behavior prevents a buildup
@@ -620,7 +620,8 @@ public class SplitLogManager implements 
     // therefore this behavior is safe.
     watcher.asyncCreate(ZKSplitLog.getRescanNode(watcher),
         TaskState.TASK_DONE.get(serverName), CreateMode.EPHEMERAL_SEQUENTIAL,
-        new CreateRescanAsyncCallback(), new Long(retries));
+        new CreateRescanAsyncCallback(), new Long(0));
+    tot_mgr_node_create_rescan_queued.incrementAndGet();
   }
 
   private void createRescanSuccess(String path) {
@@ -951,7 +952,7 @@ public class SplitLogManager implements 
             tryGetDataSetWatch(path);
           }
         }
-        createRescanNode(Long.MAX_VALUE);
+        createRescanNode();
         tot_mgr_resubmit_unassigned.incrementAndGet();
         LOG.debug("resubmitting unassigned task(s) after timeout");
       }
@@ -996,9 +997,18 @@ public class SplitLogManager implements 
     }
   }
 
+  /**
+   * @return true if there is a pending create of a task node or a RESCAN node
+   * in zookeeper. It doesn't bother about being strict - there are small
+   * windows when it can return wrong results.
+   */
   static boolean isAnyCreateZNodePending() {
-    return tot_mgr_node_create_queued.get() > tot_mgr_node_create_result.get();
+    return ((tot_mgr_node_create_queued.get() >
+             tot_mgr_node_create_result.get()) ||
+            (tot_mgr_node_create_rescan_queued.get() >
+             tot_mgr_node_create_rescan_result.get()));
   }
+
   /**
    * Asynchronous handler for zk get-data-set-watch on node results.
    * Retries on failures.
@@ -1103,6 +1113,7 @@ public class SplitLogManager implements 
 
     @Override
     public void processResult(int rc, String path, Object ctx, String name) {
+      tot_mgr_node_create_rescan_result.incrementAndGet();
       if (rc != 0) {
         if (rc == KeeperException.Code.SESSIONEXPIRED.intValue()) {
           LOG.error("ZK session expired. Master is expected to shut down. Abandoning retries.");
@@ -1111,11 +1122,8 @@ public class SplitLogManager implements 
         Long retry_count = (Long)ctx;
         LOG.warn("rc=" + KeeperException.Code.get(rc) + " for "+ path +
             " retry=" + retry_count);
-        if (retry_count == 0) {
-          createRescanFailure();
-        } else {
-          createRescanNode(retry_count - 1);
-        }
+        assert retry_count == 0;
+        createRescanFailure();
         return;
       }
       // path is the original arg, name is the actual name that was created

Modified: hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKSplitLog.java
URL: http://svn.apache.org/viewvc/hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKSplitLog.java?rev=1333200&r1=1333199&r2=1333200&view=diff
==============================================================================
--- hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKSplitLog.java (original)
+++ hbase/branches/0.89-fb/src/main/java/org/apache/hadoop/hbase/zookeeper/ZKSplitLog.java Wed May  2 20:53:40 2012
@@ -209,6 +209,10 @@ public class ZKSplitLog {
     public static AtomicLong tot_mgr_log_split_err = new AtomicLong(0);
     public static AtomicLong tot_mgr_node_create_queued = new AtomicLong(0);
     public static AtomicLong tot_mgr_node_create_result = new AtomicLong(0);
+    public static AtomicLong tot_mgr_node_create_rescan_queued =
+        new AtomicLong(0);
+    public static AtomicLong tot_mgr_node_create_rescan_result =
+        new AtomicLong(0);
     public static AtomicLong tot_mgr_node_already_exists = new AtomicLong(0);
     public static AtomicLong tot_mgr_node_create_err = new AtomicLong(0);
     public static AtomicLong tot_mgr_node_create_retry = new AtomicLong(0);