hbase-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From bus...@apache.org
Subject [38/51] [partial] hbase git commit: Published site at e73a9594c218ed969a2f5b0b356d7b8d0e1474c0.
Date Thu, 26 Nov 2015 04:30:03 GMT
http://git-wip-us.apache.org/repos/asf/hbase/blob/a986dfe6/devapidocs/src-html/org/apache/hadoop/hbase/mapreduce/TableMapReduceUtil.html
----------------------------------------------------------------------
diff --git a/devapidocs/src-html/org/apache/hadoop/hbase/mapreduce/TableMapReduceUtil.html b/devapidocs/src-html/org/apache/hadoop/hbase/mapreduce/TableMapReduceUtil.html
index 9cc08f4..88589f1 100644
--- a/devapidocs/src-html/org/apache/hadoop/hbase/mapreduce/TableMapReduceUtil.html
+++ b/devapidocs/src-html/org/apache/hadoop/hbase/mapreduce/TableMapReduceUtil.html
@@ -495,495 +495,499 @@
 <span class="sourceLineNo">487</span>        if (quorumAddress != null) {<a name="line.487"></a>
 <span class="sourceLineNo">488</span>          Configuration peerConf = HBaseConfiguration.create(job.getConfiguration());<a name="line.488"></a>
 <span class="sourceLineNo">489</span>          ZKUtil.applyClusterKeyToConf(peerConf, quorumAddress);<a name="line.489"></a>
-<span class="sourceLineNo">490</span>          Connection peerConn = ConnectionFactory.createConnection(peerConf);<a name="line.490"></a>
-<span class="sourceLineNo">491</span>          try {<a name="line.491"></a>
-<span class="sourceLineNo">492</span>            TokenUtil.addTokenForJob(peerConn, user, job);<a name="line.492"></a>
-<span class="sourceLineNo">493</span>          } finally {<a name="line.493"></a>
-<span class="sourceLineNo">494</span>            peerConn.close();<a name="line.494"></a>
-<span class="sourceLineNo">495</span>          }<a name="line.495"></a>
-<span class="sourceLineNo">496</span>        }<a name="line.496"></a>
-<span class="sourceLineNo">497</span><a name="line.497"></a>
-<span class="sourceLineNo">498</span>        Connection conn = ConnectionFactory.createConnection(job.getConfiguration());<a name="line.498"></a>
-<span class="sourceLineNo">499</span>        try {<a name="line.499"></a>
-<span class="sourceLineNo">500</span>          TokenUtil.addTokenForJob(conn, user, job);<a name="line.500"></a>
-<span class="sourceLineNo">501</span>        } finally {<a name="line.501"></a>
-<span class="sourceLineNo">502</span>          conn.close();<a name="line.502"></a>
-<span class="sourceLineNo">503</span>        }<a name="line.503"></a>
-<span class="sourceLineNo">504</span>      } catch (InterruptedException ie) {<a name="line.504"></a>
-<span class="sourceLineNo">505</span>        LOG.info("Interrupted obtaining user authentication token");<a name="line.505"></a>
-<span class="sourceLineNo">506</span>        Thread.currentThread().interrupt();<a name="line.506"></a>
-<span class="sourceLineNo">507</span>      }<a name="line.507"></a>
-<span class="sourceLineNo">508</span>    }<a name="line.508"></a>
-<span class="sourceLineNo">509</span>  }<a name="line.509"></a>
-<span class="sourceLineNo">510</span><a name="line.510"></a>
-<span class="sourceLineNo">511</span>  /**<a name="line.511"></a>
-<span class="sourceLineNo">512</span>   * Obtain an authentication token, for the specified cluster, on behalf of the current user<a name="line.512"></a>
-<span class="sourceLineNo">513</span>   * and add it to the credentials for the given map reduce job.<a name="line.513"></a>
-<span class="sourceLineNo">514</span>   *<a name="line.514"></a>
-<span class="sourceLineNo">515</span>   * The quorumAddress is the key to the ZK ensemble, which contains:<a name="line.515"></a>
-<span class="sourceLineNo">516</span>   * hbase.zookeeper.quorum, hbase.zookeeper.client.port and<a name="line.516"></a>
-<span class="sourceLineNo">517</span>   * zookeeper.znode.parent<a name="line.517"></a>
+<span class="sourceLineNo">490</span>          // apply any "hbase.mapred.output." configuration overrides<a name="line.490"></a>
+<span class="sourceLineNo">491</span>          Configuration outputOverrides =<a name="line.491"></a>
+<span class="sourceLineNo">492</span>              HBaseConfiguration.subset(peerConf, TableOutputFormat.OUTPUT_CONF_PREFIX);<a name="line.492"></a>
+<span class="sourceLineNo">493</span>          HBaseConfiguration.merge(peerConf, outputOverrides);<a name="line.493"></a>
+<span class="sourceLineNo">494</span>          Connection peerConn = ConnectionFactory.createConnection(peerConf);<a name="line.494"></a>
+<span class="sourceLineNo">495</span>          try {<a name="line.495"></a>
+<span class="sourceLineNo">496</span>            TokenUtil.addTokenForJob(peerConn, user, job);<a name="line.496"></a>
+<span class="sourceLineNo">497</span>          } finally {<a name="line.497"></a>
+<span class="sourceLineNo">498</span>            peerConn.close();<a name="line.498"></a>
+<span class="sourceLineNo">499</span>          }<a name="line.499"></a>
+<span class="sourceLineNo">500</span>        }<a name="line.500"></a>
+<span class="sourceLineNo">501</span><a name="line.501"></a>
+<span class="sourceLineNo">502</span>        Connection conn = ConnectionFactory.createConnection(job.getConfiguration());<a name="line.502"></a>
+<span class="sourceLineNo">503</span>        try {<a name="line.503"></a>
+<span class="sourceLineNo">504</span>          TokenUtil.addTokenForJob(conn, user, job);<a name="line.504"></a>
+<span class="sourceLineNo">505</span>        } finally {<a name="line.505"></a>
+<span class="sourceLineNo">506</span>          conn.close();<a name="line.506"></a>
+<span class="sourceLineNo">507</span>        }<a name="line.507"></a>
+<span class="sourceLineNo">508</span>      } catch (InterruptedException ie) {<a name="line.508"></a>
+<span class="sourceLineNo">509</span>        LOG.info("Interrupted obtaining user authentication token");<a name="line.509"></a>
+<span class="sourceLineNo">510</span>        Thread.currentThread().interrupt();<a name="line.510"></a>
+<span class="sourceLineNo">511</span>      }<a name="line.511"></a>
+<span class="sourceLineNo">512</span>    }<a name="line.512"></a>
+<span class="sourceLineNo">513</span>  }<a name="line.513"></a>
+<span class="sourceLineNo">514</span><a name="line.514"></a>
+<span class="sourceLineNo">515</span>  /**<a name="line.515"></a>
+<span class="sourceLineNo">516</span>   * Obtain an authentication token, for the specified cluster, on behalf of the current user<a name="line.516"></a>
+<span class="sourceLineNo">517</span>   * and add it to the credentials for the given map reduce job.<a name="line.517"></a>
 <span class="sourceLineNo">518</span>   *<a name="line.518"></a>
-<span class="sourceLineNo">519</span>   * @param job The job that requires the permission.<a name="line.519"></a>
-<span class="sourceLineNo">520</span>   * @param quorumAddress string that contains the 3 required configuratins<a name="line.520"></a>
-<span class="sourceLineNo">521</span>   * @throws IOException When the authentication token cannot be obtained.<a name="line.521"></a>
-<span class="sourceLineNo">522</span>   */<a name="line.522"></a>
-<span class="sourceLineNo">523</span>  public static void initCredentialsForCluster(Job job, String quorumAddress)<a name="line.523"></a>
-<span class="sourceLineNo">524</span>      throws IOException {<a name="line.524"></a>
-<span class="sourceLineNo">525</span>    UserProvider userProvider = UserProvider.instantiate(job.getConfiguration());<a name="line.525"></a>
-<span class="sourceLineNo">526</span>    if (userProvider.isHBaseSecurityEnabled()) {<a name="line.526"></a>
-<span class="sourceLineNo">527</span>      try {<a name="line.527"></a>
-<span class="sourceLineNo">528</span>        Configuration peerConf = HBaseConfiguration.create(job.getConfiguration());<a name="line.528"></a>
-<span class="sourceLineNo">529</span>        ZKUtil.applyClusterKeyToConf(peerConf, quorumAddress);<a name="line.529"></a>
-<span class="sourceLineNo">530</span>        Connection peerConn = ConnectionFactory.createConnection(peerConf);<a name="line.530"></a>
-<span class="sourceLineNo">531</span>        try {<a name="line.531"></a>
-<span class="sourceLineNo">532</span>          TokenUtil.addTokenForJob(peerConn, userProvider.getCurrent(), job);<a name="line.532"></a>
-<span class="sourceLineNo">533</span>        } finally {<a name="line.533"></a>
-<span class="sourceLineNo">534</span>          peerConn.close();<a name="line.534"></a>
-<span class="sourceLineNo">535</span>        }<a name="line.535"></a>
-<span class="sourceLineNo">536</span>      } catch (InterruptedException e) {<a name="line.536"></a>
-<span class="sourceLineNo">537</span>        LOG.info("Interrupted obtaining user authentication token");<a name="line.537"></a>
-<span class="sourceLineNo">538</span>        Thread.interrupted();<a name="line.538"></a>
-<span class="sourceLineNo">539</span>      }<a name="line.539"></a>
-<span class="sourceLineNo">540</span>    }<a name="line.540"></a>
-<span class="sourceLineNo">541</span>  }<a name="line.541"></a>
-<span class="sourceLineNo">542</span><a name="line.542"></a>
-<span class="sourceLineNo">543</span>  /**<a name="line.543"></a>
-<span class="sourceLineNo">544</span>   * Writes the given scan into a Base64 encoded string.<a name="line.544"></a>
-<span class="sourceLineNo">545</span>   *<a name="line.545"></a>
-<span class="sourceLineNo">546</span>   * @param scan  The scan to write out.<a name="line.546"></a>
-<span class="sourceLineNo">547</span>   * @return The scan saved in a Base64 encoded string.<a name="line.547"></a>
-<span class="sourceLineNo">548</span>   * @throws IOException When writing the scan fails.<a name="line.548"></a>
-<span class="sourceLineNo">549</span>   */<a name="line.549"></a>
-<span class="sourceLineNo">550</span>  static String convertScanToString(Scan scan) throws IOException {<a name="line.550"></a>
-<span class="sourceLineNo">551</span>    ClientProtos.Scan proto = ProtobufUtil.toScan(scan);<a name="line.551"></a>
-<span class="sourceLineNo">552</span>    return Base64.encodeBytes(proto.toByteArray());<a name="line.552"></a>
-<span class="sourceLineNo">553</span>  }<a name="line.553"></a>
-<span class="sourceLineNo">554</span><a name="line.554"></a>
-<span class="sourceLineNo">555</span>  /**<a name="line.555"></a>
-<span class="sourceLineNo">556</span>   * Converts the given Base64 string back into a Scan instance.<a name="line.556"></a>
-<span class="sourceLineNo">557</span>   *<a name="line.557"></a>
-<span class="sourceLineNo">558</span>   * @param base64  The scan details.<a name="line.558"></a>
-<span class="sourceLineNo">559</span>   * @return The newly created Scan instance.<a name="line.559"></a>
-<span class="sourceLineNo">560</span>   * @throws IOException When reading the scan instance fails.<a name="line.560"></a>
-<span class="sourceLineNo">561</span>   */<a name="line.561"></a>
-<span class="sourceLineNo">562</span>  static Scan convertStringToScan(String base64) throws IOException {<a name="line.562"></a>
-<span class="sourceLineNo">563</span>    byte [] decoded = Base64.decode(base64);<a name="line.563"></a>
-<span class="sourceLineNo">564</span>    ClientProtos.Scan scan;<a name="line.564"></a>
-<span class="sourceLineNo">565</span>    try {<a name="line.565"></a>
-<span class="sourceLineNo">566</span>      scan = ClientProtos.Scan.parseFrom(decoded);<a name="line.566"></a>
-<span class="sourceLineNo">567</span>    } catch (InvalidProtocolBufferException ipbe) {<a name="line.567"></a>
-<span class="sourceLineNo">568</span>      throw new IOException(ipbe);<a name="line.568"></a>
-<span class="sourceLineNo">569</span>    }<a name="line.569"></a>
-<span class="sourceLineNo">570</span><a name="line.570"></a>
-<span class="sourceLineNo">571</span>    return ProtobufUtil.toScan(scan);<a name="line.571"></a>
-<span class="sourceLineNo">572</span>  }<a name="line.572"></a>
-<span class="sourceLineNo">573</span><a name="line.573"></a>
-<span class="sourceLineNo">574</span>  /**<a name="line.574"></a>
-<span class="sourceLineNo">575</span>   * Use this before submitting a TableReduce job. It will<a name="line.575"></a>
-<span class="sourceLineNo">576</span>   * appropriately set up the JobConf.<a name="line.576"></a>
-<span class="sourceLineNo">577</span>   *<a name="line.577"></a>
-<span class="sourceLineNo">578</span>   * @param table  The output table.<a name="line.578"></a>
-<span class="sourceLineNo">579</span>   * @param reducer  The reducer class to use.<a name="line.579"></a>
-<span class="sourceLineNo">580</span>   * @param job  The current job to adjust.<a name="line.580"></a>
-<span class="sourceLineNo">581</span>   * @throws IOException When determining the region count fails.<a name="line.581"></a>
-<span class="sourceLineNo">582</span>   */<a name="line.582"></a>
-<span class="sourceLineNo">583</span>  public static void initTableReducerJob(String table,<a name="line.583"></a>
-<span class="sourceLineNo">584</span>    Class&lt;? extends TableReducer&gt; reducer, Job job)<a name="line.584"></a>
-<span class="sourceLineNo">585</span>  throws IOException {<a name="line.585"></a>
-<span class="sourceLineNo">586</span>    initTableReducerJob(table, reducer, job, null);<a name="line.586"></a>
-<span class="sourceLineNo">587</span>  }<a name="line.587"></a>
-<span class="sourceLineNo">588</span><a name="line.588"></a>
-<span class="sourceLineNo">589</span>  /**<a name="line.589"></a>
-<span class="sourceLineNo">590</span>   * Use this before submitting a TableReduce job. It will<a name="line.590"></a>
-<span class="sourceLineNo">591</span>   * appropriately set up the JobConf.<a name="line.591"></a>
-<span class="sourceLineNo">592</span>   *<a name="line.592"></a>
-<span class="sourceLineNo">593</span>   * @param table  The output table.<a name="line.593"></a>
-<span class="sourceLineNo">594</span>   * @param reducer  The reducer class to use.<a name="line.594"></a>
-<span class="sourceLineNo">595</span>   * @param job  The current job to adjust.<a name="line.595"></a>
-<span class="sourceLineNo">596</span>   * @param partitioner  Partitioner to use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.596"></a>
-<span class="sourceLineNo">597</span>   * default partitioner.<a name="line.597"></a>
-<span class="sourceLineNo">598</span>   * @throws IOException When determining the region count fails.<a name="line.598"></a>
-<span class="sourceLineNo">599</span>   */<a name="line.599"></a>
-<span class="sourceLineNo">600</span>  public static void initTableReducerJob(String table,<a name="line.600"></a>
-<span class="sourceLineNo">601</span>    Class&lt;? extends TableReducer&gt; reducer, Job job,<a name="line.601"></a>
-<span class="sourceLineNo">602</span>    Class partitioner) throws IOException {<a name="line.602"></a>
-<span class="sourceLineNo">603</span>    initTableReducerJob(table, reducer, job, partitioner, null, null, null);<a name="line.603"></a>
-<span class="sourceLineNo">604</span>  }<a name="line.604"></a>
-<span class="sourceLineNo">605</span><a name="line.605"></a>
-<span class="sourceLineNo">606</span>  /**<a name="line.606"></a>
-<span class="sourceLineNo">607</span>   * Use this before submitting a TableReduce job. It will<a name="line.607"></a>
-<span class="sourceLineNo">608</span>   * appropriately set up the JobConf.<a name="line.608"></a>
-<span class="sourceLineNo">609</span>   *<a name="line.609"></a>
-<span class="sourceLineNo">610</span>   * @param table  The output table.<a name="line.610"></a>
-<span class="sourceLineNo">611</span>   * @param reducer  The reducer class to use.<a name="line.611"></a>
-<span class="sourceLineNo">612</span>   * @param job  The current job to adjust.  Make sure the passed job is<a name="line.612"></a>
-<span class="sourceLineNo">613</span>   * carrying all necessary HBase configuration.<a name="line.613"></a>
-<span class="sourceLineNo">614</span>   * @param partitioner  Partitioner to use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.614"></a>
-<span class="sourceLineNo">615</span>   * default partitioner.<a name="line.615"></a>
-<span class="sourceLineNo">616</span>   * @param quorumAddress Distant cluster to write to; default is null for<a name="line.616"></a>
-<span class="sourceLineNo">617</span>   * output to the cluster that is designated in &lt;code&gt;hbase-site.xml&lt;/code&gt;.<a name="line.617"></a>
-<span class="sourceLineNo">618</span>   * Set this String to the zookeeper ensemble of an alternate remote cluster<a name="line.618"></a>
-<span class="sourceLineNo">619</span>   * when you would have the reduce write a cluster that is other than the<a name="line.619"></a>
-<span class="sourceLineNo">620</span>   * default; e.g. copying tables between clusters, the source would be<a name="line.620"></a>
-<span class="sourceLineNo">621</span>   * designated by &lt;code&gt;hbase-site.xml&lt;/code&gt; and this param would have the<a name="line.621"></a>
-<span class="sourceLineNo">622</span>   * ensemble address of the remote cluster.  The format to pass is particular.<a name="line.622"></a>
-<span class="sourceLineNo">623</span>   * Pass &lt;code&gt; &amp;lt;hbase.zookeeper.quorum&amp;gt;:&amp;lt;<a name="line.623"></a>
-<span class="sourceLineNo">624</span>   *             hbase.zookeeper.client.port&amp;gt;:&amp;lt;zookeeper.znode.parent&amp;gt;<a name="line.624"></a>
-<span class="sourceLineNo">625</span>   * &lt;/code&gt; such as &lt;code&gt;server,server2,server3:2181:/hbase&lt;/code&gt;.<a name="line.625"></a>
-<span class="sourceLineNo">626</span>   * @param serverClass redefined hbase.regionserver.class<a name="line.626"></a>
-<span class="sourceLineNo">627</span>   * @param serverImpl redefined hbase.regionserver.impl<a name="line.627"></a>
-<span class="sourceLineNo">628</span>   * @throws IOException When determining the region count fails.<a name="line.628"></a>
-<span class="sourceLineNo">629</span>   */<a name="line.629"></a>
-<span class="sourceLineNo">630</span>  public static void initTableReducerJob(String table,<a name="line.630"></a>
-<span class="sourceLineNo">631</span>    Class&lt;? extends TableReducer&gt; reducer, Job job,<a name="line.631"></a>
-<span class="sourceLineNo">632</span>    Class partitioner, String quorumAddress, String serverClass,<a name="line.632"></a>
-<span class="sourceLineNo">633</span>    String serverImpl) throws IOException {<a name="line.633"></a>
-<span class="sourceLineNo">634</span>    initTableReducerJob(table, reducer, job, partitioner, quorumAddress,<a name="line.634"></a>
-<span class="sourceLineNo">635</span>        serverClass, serverImpl, true);<a name="line.635"></a>
-<span class="sourceLineNo">636</span>  }<a name="line.636"></a>
-<span class="sourceLineNo">637</span><a name="line.637"></a>
-<span class="sourceLineNo">638</span>  /**<a name="line.638"></a>
-<span class="sourceLineNo">639</span>   * Use this before submitting a TableReduce job. It will<a name="line.639"></a>
-<span class="sourceLineNo">640</span>   * appropriately set up the JobConf.<a name="line.640"></a>
-<span class="sourceLineNo">641</span>   *<a name="line.641"></a>
-<span class="sourceLineNo">642</span>   * @param table  The output table.<a name="line.642"></a>
-<span class="sourceLineNo">643</span>   * @param reducer  The reducer class to use.<a name="line.643"></a>
-<span class="sourceLineNo">644</span>   * @param job  The current job to adjust.  Make sure the passed job is<a name="line.644"></a>
-<span class="sourceLineNo">645</span>   * carrying all necessary HBase configuration.<a name="line.645"></a>
-<span class="sourceLineNo">646</span>   * @param partitioner  Partitioner to use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.646"></a>
-<span class="sourceLineNo">647</span>   * default partitioner.<a name="line.647"></a>
-<span class="sourceLineNo">648</span>   * @param quorumAddress Distant cluster to write to; default is null for<a name="line.648"></a>
-<span class="sourceLineNo">649</span>   * output to the cluster that is designated in &lt;code&gt;hbase-site.xml&lt;/code&gt;.<a name="line.649"></a>
-<span class="sourceLineNo">650</span>   * Set this String to the zookeeper ensemble of an alternate remote cluster<a name="line.650"></a>
-<span class="sourceLineNo">651</span>   * when you would have the reduce write a cluster that is other than the<a name="line.651"></a>
-<span class="sourceLineNo">652</span>   * default; e.g. copying tables between clusters, the source would be<a name="line.652"></a>
-<span class="sourceLineNo">653</span>   * designated by &lt;code&gt;hbase-site.xml&lt;/code&gt; and this param would have the<a name="line.653"></a>
-<span class="sourceLineNo">654</span>   * ensemble address of the remote cluster.  The format to pass is particular.<a name="line.654"></a>
-<span class="sourceLineNo">655</span>   * Pass &lt;code&gt; &amp;lt;hbase.zookeeper.quorum&amp;gt;:&amp;lt;<a name="line.655"></a>
-<span class="sourceLineNo">656</span>   *             hbase.zookeeper.client.port&amp;gt;:&amp;lt;zookeeper.znode.parent&amp;gt;<a name="line.656"></a>
-<span class="sourceLineNo">657</span>   * &lt;/code&gt; such as &lt;code&gt;server,server2,server3:2181:/hbase&lt;/code&gt;.<a name="line.657"></a>
-<span class="sourceLineNo">658</span>   * @param serverClass redefined hbase.regionserver.class<a name="line.658"></a>
-<span class="sourceLineNo">659</span>   * @param serverImpl redefined hbase.regionserver.impl<a name="line.659"></a>
-<span class="sourceLineNo">660</span>   * @param addDependencyJars upload HBase jars and jars for any of the configured<a name="line.660"></a>
-<span class="sourceLineNo">661</span>   *           job classes via the distributed cache (tmpjars).<a name="line.661"></a>
-<span class="sourceLineNo">662</span>   * @throws IOException When determining the region count fails.<a name="line.662"></a>
-<span class="sourceLineNo">663</span>   */<a name="line.663"></a>
-<span class="sourceLineNo">664</span>  public static void initTableReducerJob(String table,<a name="line.664"></a>
-<span class="sourceLineNo">665</span>    Class&lt;? extends TableReducer&gt; reducer, Job job,<a name="line.665"></a>
-<span class="sourceLineNo">666</span>    Class partitioner, String quorumAddress, String serverClass,<a name="line.666"></a>
-<span class="sourceLineNo">667</span>    String serverImpl, boolean addDependencyJars) throws IOException {<a name="line.667"></a>
-<span class="sourceLineNo">668</span><a name="line.668"></a>
-<span class="sourceLineNo">669</span>    Configuration conf = job.getConfiguration();<a name="line.669"></a>
-<span class="sourceLineNo">670</span>    HBaseConfiguration.merge(conf, HBaseConfiguration.create(conf));<a name="line.670"></a>
-<span class="sourceLineNo">671</span>    job.setOutputFormatClass(TableOutputFormat.class);<a name="line.671"></a>
-<span class="sourceLineNo">672</span>    if (reducer != null) job.setReducerClass(reducer);<a name="line.672"></a>
-<span class="sourceLineNo">673</span>    conf.set(TableOutputFormat.OUTPUT_TABLE, table);<a name="line.673"></a>
-<span class="sourceLineNo">674</span>    conf.setStrings("io.serializations", conf.get("io.serializations"),<a name="line.674"></a>
-<span class="sourceLineNo">675</span>        MutationSerialization.class.getName(), ResultSerialization.class.getName());<a name="line.675"></a>
-<span class="sourceLineNo">676</span>    // If passed a quorum/ensemble address, pass it on to TableOutputFormat.<a name="line.676"></a>
-<span class="sourceLineNo">677</span>    if (quorumAddress != null) {<a name="line.677"></a>
-<span class="sourceLineNo">678</span>      // Calling this will validate the format<a name="line.678"></a>
-<span class="sourceLineNo">679</span>      ZKUtil.transformClusterKey(quorumAddress);<a name="line.679"></a>
-<span class="sourceLineNo">680</span>      conf.set(TableOutputFormat.QUORUM_ADDRESS,quorumAddress);<a name="line.680"></a>
-<span class="sourceLineNo">681</span>    }<a name="line.681"></a>
-<span class="sourceLineNo">682</span>    if (serverClass != null &amp;&amp; serverImpl != null) {<a name="line.682"></a>
-<span class="sourceLineNo">683</span>      conf.set(TableOutputFormat.REGION_SERVER_CLASS, serverClass);<a name="line.683"></a>
-<span class="sourceLineNo">684</span>      conf.set(TableOutputFormat.REGION_SERVER_IMPL, serverImpl);<a name="line.684"></a>
+<span class="sourceLineNo">519</span>   * The quorumAddress is the key to the ZK ensemble, which contains:<a name="line.519"></a>
+<span class="sourceLineNo">520</span>   * hbase.zookeeper.quorum, hbase.zookeeper.client.port and<a name="line.520"></a>
+<span class="sourceLineNo">521</span>   * zookeeper.znode.parent<a name="line.521"></a>
+<span class="sourceLineNo">522</span>   *<a name="line.522"></a>
+<span class="sourceLineNo">523</span>   * @param job The job that requires the permission.<a name="line.523"></a>
+<span class="sourceLineNo">524</span>   * @param quorumAddress string that contains the 3 required configuratins<a name="line.524"></a>
+<span class="sourceLineNo">525</span>   * @throws IOException When the authentication token cannot be obtained.<a name="line.525"></a>
+<span class="sourceLineNo">526</span>   */<a name="line.526"></a>
+<span class="sourceLineNo">527</span>  public static void initCredentialsForCluster(Job job, String quorumAddress)<a name="line.527"></a>
+<span class="sourceLineNo">528</span>      throws IOException {<a name="line.528"></a>
+<span class="sourceLineNo">529</span>    UserProvider userProvider = UserProvider.instantiate(job.getConfiguration());<a name="line.529"></a>
+<span class="sourceLineNo">530</span>    if (userProvider.isHBaseSecurityEnabled()) {<a name="line.530"></a>
+<span class="sourceLineNo">531</span>      try {<a name="line.531"></a>
+<span class="sourceLineNo">532</span>        Configuration peerConf = HBaseConfiguration.create(job.getConfiguration());<a name="line.532"></a>
+<span class="sourceLineNo">533</span>        ZKUtil.applyClusterKeyToConf(peerConf, quorumAddress);<a name="line.533"></a>
+<span class="sourceLineNo">534</span>        Connection peerConn = ConnectionFactory.createConnection(peerConf);<a name="line.534"></a>
+<span class="sourceLineNo">535</span>        try {<a name="line.535"></a>
+<span class="sourceLineNo">536</span>          TokenUtil.addTokenForJob(peerConn, userProvider.getCurrent(), job);<a name="line.536"></a>
+<span class="sourceLineNo">537</span>        } finally {<a name="line.537"></a>
+<span class="sourceLineNo">538</span>          peerConn.close();<a name="line.538"></a>
+<span class="sourceLineNo">539</span>        }<a name="line.539"></a>
+<span class="sourceLineNo">540</span>      } catch (InterruptedException e) {<a name="line.540"></a>
+<span class="sourceLineNo">541</span>        LOG.info("Interrupted obtaining user authentication token");<a name="line.541"></a>
+<span class="sourceLineNo">542</span>        Thread.interrupted();<a name="line.542"></a>
+<span class="sourceLineNo">543</span>      }<a name="line.543"></a>
+<span class="sourceLineNo">544</span>    }<a name="line.544"></a>
+<span class="sourceLineNo">545</span>  }<a name="line.545"></a>
+<span class="sourceLineNo">546</span><a name="line.546"></a>
+<span class="sourceLineNo">547</span>  /**<a name="line.547"></a>
+<span class="sourceLineNo">548</span>   * Writes the given scan into a Base64 encoded string.<a name="line.548"></a>
+<span class="sourceLineNo">549</span>   *<a name="line.549"></a>
+<span class="sourceLineNo">550</span>   * @param scan  The scan to write out.<a name="line.550"></a>
+<span class="sourceLineNo">551</span>   * @return The scan saved in a Base64 encoded string.<a name="line.551"></a>
+<span class="sourceLineNo">552</span>   * @throws IOException When writing the scan fails.<a name="line.552"></a>
+<span class="sourceLineNo">553</span>   */<a name="line.553"></a>
+<span class="sourceLineNo">554</span>  static String convertScanToString(Scan scan) throws IOException {<a name="line.554"></a>
+<span class="sourceLineNo">555</span>    ClientProtos.Scan proto = ProtobufUtil.toScan(scan);<a name="line.555"></a>
+<span class="sourceLineNo">556</span>    return Base64.encodeBytes(proto.toByteArray());<a name="line.556"></a>
+<span class="sourceLineNo">557</span>  }<a name="line.557"></a>
+<span class="sourceLineNo">558</span><a name="line.558"></a>
+<span class="sourceLineNo">559</span>  /**<a name="line.559"></a>
+<span class="sourceLineNo">560</span>   * Converts the given Base64 string back into a Scan instance.<a name="line.560"></a>
+<span class="sourceLineNo">561</span>   *<a name="line.561"></a>
+<span class="sourceLineNo">562</span>   * @param base64  The scan details.<a name="line.562"></a>
+<span class="sourceLineNo">563</span>   * @return The newly created Scan instance.<a name="line.563"></a>
+<span class="sourceLineNo">564</span>   * @throws IOException When reading the scan instance fails.<a name="line.564"></a>
+<span class="sourceLineNo">565</span>   */<a name="line.565"></a>
+<span class="sourceLineNo">566</span>  static Scan convertStringToScan(String base64) throws IOException {<a name="line.566"></a>
+<span class="sourceLineNo">567</span>    byte [] decoded = Base64.decode(base64);<a name="line.567"></a>
+<span class="sourceLineNo">568</span>    ClientProtos.Scan scan;<a name="line.568"></a>
+<span class="sourceLineNo">569</span>    try {<a name="line.569"></a>
+<span class="sourceLineNo">570</span>      scan = ClientProtos.Scan.parseFrom(decoded);<a name="line.570"></a>
+<span class="sourceLineNo">571</span>    } catch (InvalidProtocolBufferException ipbe) {<a name="line.571"></a>
+<span class="sourceLineNo">572</span>      throw new IOException(ipbe);<a name="line.572"></a>
+<span class="sourceLineNo">573</span>    }<a name="line.573"></a>
+<span class="sourceLineNo">574</span><a name="line.574"></a>
+<span class="sourceLineNo">575</span>    return ProtobufUtil.toScan(scan);<a name="line.575"></a>
+<span class="sourceLineNo">576</span>  }<a name="line.576"></a>
+<span class="sourceLineNo">577</span><a name="line.577"></a>
+<span class="sourceLineNo">578</span>  /**<a name="line.578"></a>
+<span class="sourceLineNo">579</span>   * Use this before submitting a TableReduce job. It will<a name="line.579"></a>
+<span class="sourceLineNo">580</span>   * appropriately set up the JobConf.<a name="line.580"></a>
+<span class="sourceLineNo">581</span>   *<a name="line.581"></a>
+<span class="sourceLineNo">582</span>   * @param table  The output table.<a name="line.582"></a>
+<span class="sourceLineNo">583</span>   * @param reducer  The reducer class to use.<a name="line.583"></a>
+<span class="sourceLineNo">584</span>   * @param job  The current job to adjust.<a name="line.584"></a>
+<span class="sourceLineNo">585</span>   * @throws IOException When determining the region count fails.<a name="line.585"></a>
+<span class="sourceLineNo">586</span>   */<a name="line.586"></a>
+<span class="sourceLineNo">587</span>  public static void initTableReducerJob(String table,<a name="line.587"></a>
+<span class="sourceLineNo">588</span>    Class&lt;? extends TableReducer&gt; reducer, Job job)<a name="line.588"></a>
+<span class="sourceLineNo">589</span>  throws IOException {<a name="line.589"></a>
+<span class="sourceLineNo">590</span>    initTableReducerJob(table, reducer, job, null);<a name="line.590"></a>
+<span class="sourceLineNo">591</span>  }<a name="line.591"></a>
+<span class="sourceLineNo">592</span><a name="line.592"></a>
+<span class="sourceLineNo">593</span>  /**<a name="line.593"></a>
+<span class="sourceLineNo">594</span>   * Use this before submitting a TableReduce job. It will<a name="line.594"></a>
+<span class="sourceLineNo">595</span>   * appropriately set up the JobConf.<a name="line.595"></a>
+<span class="sourceLineNo">596</span>   *<a name="line.596"></a>
+<span class="sourceLineNo">597</span>   * @param table  The output table.<a name="line.597"></a>
+<span class="sourceLineNo">598</span>   * @param reducer  The reducer class to use.<a name="line.598"></a>
+<span class="sourceLineNo">599</span>   * @param job  The current job to adjust.<a name="line.599"></a>
+<span class="sourceLineNo">600</span>   * @param partitioner  Partitioner to use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.600"></a>
+<span class="sourceLineNo">601</span>   * default partitioner.<a name="line.601"></a>
+<span class="sourceLineNo">602</span>   * @throws IOException When determining the region count fails.<a name="line.602"></a>
+<span class="sourceLineNo">603</span>   */<a name="line.603"></a>
+<span class="sourceLineNo">604</span>  public static void initTableReducerJob(String table,<a name="line.604"></a>
+<span class="sourceLineNo">605</span>    Class&lt;? extends TableReducer&gt; reducer, Job job,<a name="line.605"></a>
+<span class="sourceLineNo">606</span>    Class partitioner) throws IOException {<a name="line.606"></a>
+<span class="sourceLineNo">607</span>    initTableReducerJob(table, reducer, job, partitioner, null, null, null);<a name="line.607"></a>
+<span class="sourceLineNo">608</span>  }<a name="line.608"></a>
+<span class="sourceLineNo">609</span><a name="line.609"></a>
+<span class="sourceLineNo">610</span>  /**<a name="line.610"></a>
+<span class="sourceLineNo">611</span>   * Use this before submitting a TableReduce job. It will<a name="line.611"></a>
+<span class="sourceLineNo">612</span>   * appropriately set up the JobConf.<a name="line.612"></a>
+<span class="sourceLineNo">613</span>   *<a name="line.613"></a>
+<span class="sourceLineNo">614</span>   * @param table  The output table.<a name="line.614"></a>
+<span class="sourceLineNo">615</span>   * @param reducer  The reducer class to use.<a name="line.615"></a>
+<span class="sourceLineNo">616</span>   * @param job  The current job to adjust.  Make sure the passed job is<a name="line.616"></a>
+<span class="sourceLineNo">617</span>   * carrying all necessary HBase configuration.<a name="line.617"></a>
+<span class="sourceLineNo">618</span>   * @param partitioner  Partitioner to use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.618"></a>
+<span class="sourceLineNo">619</span>   * default partitioner.<a name="line.619"></a>
+<span class="sourceLineNo">620</span>   * @param quorumAddress Distant cluster to write to; default is null for<a name="line.620"></a>
+<span class="sourceLineNo">621</span>   * output to the cluster that is designated in &lt;code&gt;hbase-site.xml&lt;/code&gt;.<a name="line.621"></a>
+<span class="sourceLineNo">622</span>   * Set this String to the zookeeper ensemble of an alternate remote cluster<a name="line.622"></a>
+<span class="sourceLineNo">623</span>   * when you would have the reduce write a cluster that is other than the<a name="line.623"></a>
+<span class="sourceLineNo">624</span>   * default; e.g. copying tables between clusters, the source would be<a name="line.624"></a>
+<span class="sourceLineNo">625</span>   * designated by &lt;code&gt;hbase-site.xml&lt;/code&gt; and this param would have the<a name="line.625"></a>
+<span class="sourceLineNo">626</span>   * ensemble address of the remote cluster.  The format to pass is particular.<a name="line.626"></a>
+<span class="sourceLineNo">627</span>   * Pass &lt;code&gt; &amp;lt;hbase.zookeeper.quorum&amp;gt;:&amp;lt;<a name="line.627"></a>
+<span class="sourceLineNo">628</span>   *             hbase.zookeeper.client.port&amp;gt;:&amp;lt;zookeeper.znode.parent&amp;gt;<a name="line.628"></a>
+<span class="sourceLineNo">629</span>   * &lt;/code&gt; such as &lt;code&gt;server,server2,server3:2181:/hbase&lt;/code&gt;.<a name="line.629"></a>
+<span class="sourceLineNo">630</span>   * @param serverClass redefined hbase.regionserver.class<a name="line.630"></a>
+<span class="sourceLineNo">631</span>   * @param serverImpl redefined hbase.regionserver.impl<a name="line.631"></a>
+<span class="sourceLineNo">632</span>   * @throws IOException When determining the region count fails.<a name="line.632"></a>
+<span class="sourceLineNo">633</span>   */<a name="line.633"></a>
+<span class="sourceLineNo">634</span>  public static void initTableReducerJob(String table,<a name="line.634"></a>
+<span class="sourceLineNo">635</span>    Class&lt;? extends TableReducer&gt; reducer, Job job,<a name="line.635"></a>
+<span class="sourceLineNo">636</span>    Class partitioner, String quorumAddress, String serverClass,<a name="line.636"></a>
+<span class="sourceLineNo">637</span>    String serverImpl) throws IOException {<a name="line.637"></a>
+<span class="sourceLineNo">638</span>    initTableReducerJob(table, reducer, job, partitioner, quorumAddress,<a name="line.638"></a>
+<span class="sourceLineNo">639</span>        serverClass, serverImpl, true);<a name="line.639"></a>
+<span class="sourceLineNo">640</span>  }<a name="line.640"></a>
+<span class="sourceLineNo">641</span><a name="line.641"></a>
+<span class="sourceLineNo">642</span>  /**<a name="line.642"></a>
+<span class="sourceLineNo">643</span>   * Use this before submitting a TableReduce job. It will<a name="line.643"></a>
+<span class="sourceLineNo">644</span>   * appropriately set up the JobConf.<a name="line.644"></a>
+<span class="sourceLineNo">645</span>   *<a name="line.645"></a>
+<span class="sourceLineNo">646</span>   * @param table  The output table.<a name="line.646"></a>
+<span class="sourceLineNo">647</span>   * @param reducer  The reducer class to use.<a name="line.647"></a>
+<span class="sourceLineNo">648</span>   * @param job  The current job to adjust.  Make sure the passed job is<a name="line.648"></a>
+<span class="sourceLineNo">649</span>   * carrying all necessary HBase configuration.<a name="line.649"></a>
+<span class="sourceLineNo">650</span>   * @param partitioner  Partitioner to use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.650"></a>
+<span class="sourceLineNo">651</span>   * default partitioner.<a name="line.651"></a>
+<span class="sourceLineNo">652</span>   * @param quorumAddress Distant cluster to write to; default is null for<a name="line.652"></a>
+<span class="sourceLineNo">653</span>   * output to the cluster that is designated in &lt;code&gt;hbase-site.xml&lt;/code&gt;.<a name="line.653"></a>
+<span class="sourceLineNo">654</span>   * Set this String to the zookeeper ensemble of an alternate remote cluster<a name="line.654"></a>
+<span class="sourceLineNo">655</span>   * when you would have the reduce write a cluster that is other than the<a name="line.655"></a>
+<span class="sourceLineNo">656</span>   * default; e.g. copying tables between clusters, the source would be<a name="line.656"></a>
+<span class="sourceLineNo">657</span>   * designated by &lt;code&gt;hbase-site.xml&lt;/code&gt; and this param would have the<a name="line.657"></a>
+<span class="sourceLineNo">658</span>   * ensemble address of the remote cluster.  The format to pass is particular.<a name="line.658"></a>
+<span class="sourceLineNo">659</span>   * Pass &lt;code&gt; &amp;lt;hbase.zookeeper.quorum&amp;gt;:&amp;lt;<a name="line.659"></a>
+<span class="sourceLineNo">660</span>   *             hbase.zookeeper.client.port&amp;gt;:&amp;lt;zookeeper.znode.parent&amp;gt;<a name="line.660"></a>
+<span class="sourceLineNo">661</span>   * &lt;/code&gt; such as &lt;code&gt;server,server2,server3:2181:/hbase&lt;/code&gt;.<a name="line.661"></a>
+<span class="sourceLineNo">662</span>   * @param serverClass redefined hbase.regionserver.class<a name="line.662"></a>
+<span class="sourceLineNo">663</span>   * @param serverImpl redefined hbase.regionserver.impl<a name="line.663"></a>
+<span class="sourceLineNo">664</span>   * @param addDependencyJars upload HBase jars and jars for any of the configured<a name="line.664"></a>
+<span class="sourceLineNo">665</span>   *           job classes via the distributed cache (tmpjars).<a name="line.665"></a>
+<span class="sourceLineNo">666</span>   * @throws IOException When determining the region count fails.<a name="line.666"></a>
+<span class="sourceLineNo">667</span>   */<a name="line.667"></a>
+<span class="sourceLineNo">668</span>  public static void initTableReducerJob(String table,<a name="line.668"></a>
+<span class="sourceLineNo">669</span>    Class&lt;? extends TableReducer&gt; reducer, Job job,<a name="line.669"></a>
+<span class="sourceLineNo">670</span>    Class partitioner, String quorumAddress, String serverClass,<a name="line.670"></a>
+<span class="sourceLineNo">671</span>    String serverImpl, boolean addDependencyJars) throws IOException {<a name="line.671"></a>
+<span class="sourceLineNo">672</span><a name="line.672"></a>
+<span class="sourceLineNo">673</span>    Configuration conf = job.getConfiguration();<a name="line.673"></a>
+<span class="sourceLineNo">674</span>    HBaseConfiguration.merge(conf, HBaseConfiguration.create(conf));<a name="line.674"></a>
+<span class="sourceLineNo">675</span>    job.setOutputFormatClass(TableOutputFormat.class);<a name="line.675"></a>
+<span class="sourceLineNo">676</span>    if (reducer != null) job.setReducerClass(reducer);<a name="line.676"></a>
+<span class="sourceLineNo">677</span>    conf.set(TableOutputFormat.OUTPUT_TABLE, table);<a name="line.677"></a>
+<span class="sourceLineNo">678</span>    conf.setStrings("io.serializations", conf.get("io.serializations"),<a name="line.678"></a>
+<span class="sourceLineNo">679</span>        MutationSerialization.class.getName(), ResultSerialization.class.getName());<a name="line.679"></a>
+<span class="sourceLineNo">680</span>    // If passed a quorum/ensemble address, pass it on to TableOutputFormat.<a name="line.680"></a>
+<span class="sourceLineNo">681</span>    if (quorumAddress != null) {<a name="line.681"></a>
+<span class="sourceLineNo">682</span>      // Calling this will validate the format<a name="line.682"></a>
+<span class="sourceLineNo">683</span>      ZKUtil.transformClusterKey(quorumAddress);<a name="line.683"></a>
+<span class="sourceLineNo">684</span>      conf.set(TableOutputFormat.QUORUM_ADDRESS,quorumAddress);<a name="line.684"></a>
 <span class="sourceLineNo">685</span>    }<a name="line.685"></a>
-<span class="sourceLineNo">686</span>    job.setOutputKeyClass(ImmutableBytesWritable.class);<a name="line.686"></a>
-<span class="sourceLineNo">687</span>    job.setOutputValueClass(Writable.class);<a name="line.687"></a>
-<span class="sourceLineNo">688</span>    if (partitioner == HRegionPartitioner.class) {<a name="line.688"></a>
-<span class="sourceLineNo">689</span>      job.setPartitionerClass(HRegionPartitioner.class);<a name="line.689"></a>
-<span class="sourceLineNo">690</span>      int regions = MetaTableAccessor.getRegionCount(conf, TableName.valueOf(table));<a name="line.690"></a>
-<span class="sourceLineNo">691</span>      if (job.getNumReduceTasks() &gt; regions) {<a name="line.691"></a>
-<span class="sourceLineNo">692</span>        job.setNumReduceTasks(regions);<a name="line.692"></a>
-<span class="sourceLineNo">693</span>      }<a name="line.693"></a>
-<span class="sourceLineNo">694</span>    } else if (partitioner != null) {<a name="line.694"></a>
-<span class="sourceLineNo">695</span>      job.setPartitionerClass(partitioner);<a name="line.695"></a>
-<span class="sourceLineNo">696</span>    }<a name="line.696"></a>
-<span class="sourceLineNo">697</span><a name="line.697"></a>
-<span class="sourceLineNo">698</span>    if (addDependencyJars) {<a name="line.698"></a>
-<span class="sourceLineNo">699</span>      addDependencyJars(job);<a name="line.699"></a>
+<span class="sourceLineNo">686</span>    if (serverClass != null &amp;&amp; serverImpl != null) {<a name="line.686"></a>
+<span class="sourceLineNo">687</span>      conf.set(TableOutputFormat.REGION_SERVER_CLASS, serverClass);<a name="line.687"></a>
+<span class="sourceLineNo">688</span>      conf.set(TableOutputFormat.REGION_SERVER_IMPL, serverImpl);<a name="line.688"></a>
+<span class="sourceLineNo">689</span>    }<a name="line.689"></a>
+<span class="sourceLineNo">690</span>    job.setOutputKeyClass(ImmutableBytesWritable.class);<a name="line.690"></a>
+<span class="sourceLineNo">691</span>    job.setOutputValueClass(Writable.class);<a name="line.691"></a>
+<span class="sourceLineNo">692</span>    if (partitioner == HRegionPartitioner.class) {<a name="line.692"></a>
+<span class="sourceLineNo">693</span>      job.setPartitionerClass(HRegionPartitioner.class);<a name="line.693"></a>
+<span class="sourceLineNo">694</span>      int regions = MetaTableAccessor.getRegionCount(conf, TableName.valueOf(table));<a name="line.694"></a>
+<span class="sourceLineNo">695</span>      if (job.getNumReduceTasks() &gt; regions) {<a name="line.695"></a>
+<span class="sourceLineNo">696</span>        job.setNumReduceTasks(regions);<a name="line.696"></a>
+<span class="sourceLineNo">697</span>      }<a name="line.697"></a>
+<span class="sourceLineNo">698</span>    } else if (partitioner != null) {<a name="line.698"></a>
+<span class="sourceLineNo">699</span>      job.setPartitionerClass(partitioner);<a name="line.699"></a>
 <span class="sourceLineNo">700</span>    }<a name="line.700"></a>
 <span class="sourceLineNo">701</span><a name="line.701"></a>
-<span class="sourceLineNo">702</span>    initCredentials(job);<a name="line.702"></a>
-<span class="sourceLineNo">703</span>  }<a name="line.703"></a>
-<span class="sourceLineNo">704</span><a name="line.704"></a>
-<span class="sourceLineNo">705</span>  /**<a name="line.705"></a>
-<span class="sourceLineNo">706</span>   * Ensures that the given number of reduce tasks for the given job<a name="line.706"></a>
-<span class="sourceLineNo">707</span>   * configuration does not exceed the number of regions for the given table.<a name="line.707"></a>
-<span class="sourceLineNo">708</span>   *<a name="line.708"></a>
-<span class="sourceLineNo">709</span>   * @param table  The table to get the region count for.<a name="line.709"></a>
-<span class="sourceLineNo">710</span>   * @param job  The current job to adjust.<a name="line.710"></a>
-<span class="sourceLineNo">711</span>   * @throws IOException When retrieving the table details fails.<a name="line.711"></a>
-<span class="sourceLineNo">712</span>   */<a name="line.712"></a>
-<span class="sourceLineNo">713</span>  public static void limitNumReduceTasks(String table, Job job)<a name="line.713"></a>
-<span class="sourceLineNo">714</span>  throws IOException {<a name="line.714"></a>
-<span class="sourceLineNo">715</span>    int regions =<a name="line.715"></a>
-<span class="sourceLineNo">716</span>      MetaTableAccessor.getRegionCount(job.getConfiguration(), TableName.valueOf(table));<a name="line.716"></a>
-<span class="sourceLineNo">717</span>    if (job.getNumReduceTasks() &gt; regions)<a name="line.717"></a>
-<span class="sourceLineNo">718</span>      job.setNumReduceTasks(regions);<a name="line.718"></a>
-<span class="sourceLineNo">719</span>  }<a name="line.719"></a>
-<span class="sourceLineNo">720</span><a name="line.720"></a>
-<span class="sourceLineNo">721</span>  /**<a name="line.721"></a>
-<span class="sourceLineNo">722</span>   * Sets the number of reduce tasks for the given job configuration to the<a name="line.722"></a>
-<span class="sourceLineNo">723</span>   * number of regions the given table has.<a name="line.723"></a>
-<span class="sourceLineNo">724</span>   *<a name="line.724"></a>
-<span class="sourceLineNo">725</span>   * @param table  The table to get the region count for.<a name="line.725"></a>
-<span class="sourceLineNo">726</span>   * @param job  The current job to adjust.<a name="line.726"></a>
-<span class="sourceLineNo">727</span>   * @throws IOException When retrieving the table details fails.<a name="line.727"></a>
-<span class="sourceLineNo">728</span>   */<a name="line.728"></a>
-<span class="sourceLineNo">729</span>  public static void setNumReduceTasks(String table, Job job)<a name="line.729"></a>
-<span class="sourceLineNo">730</span>  throws IOException {<a name="line.730"></a>
-<span class="sourceLineNo">731</span>    job.setNumReduceTasks(MetaTableAccessor.getRegionCount(job.getConfiguration(),<a name="line.731"></a>
-<span class="sourceLineNo">732</span>       TableName.valueOf(table)));<a name="line.732"></a>
-<span class="sourceLineNo">733</span>  }<a name="line.733"></a>
-<span class="sourceLineNo">734</span><a name="line.734"></a>
-<span class="sourceLineNo">735</span>  /**<a name="line.735"></a>
-<span class="sourceLineNo">736</span>   * Sets the number of rows to return and cache with each scanner iteration.<a name="line.736"></a>
-<span class="sourceLineNo">737</span>   * Higher caching values will enable faster mapreduce jobs at the expense of<a name="line.737"></a>
-<span class="sourceLineNo">738</span>   * requiring more heap to contain the cached rows.<a name="line.738"></a>
-<span class="sourceLineNo">739</span>   *<a name="line.739"></a>
-<span class="sourceLineNo">740</span>   * @param job The current job to adjust.<a name="line.740"></a>
-<span class="sourceLineNo">741</span>   * @param batchSize The number of rows to return in batch with each scanner<a name="line.741"></a>
-<span class="sourceLineNo">742</span>   * iteration.<a name="line.742"></a>
-<span class="sourceLineNo">743</span>   */<a name="line.743"></a>
-<span class="sourceLineNo">744</span>  public static void setScannerCaching(Job job, int batchSize) {<a name="line.744"></a>
-<span class="sourceLineNo">745</span>    job.getConfiguration().setInt("hbase.client.scanner.caching", batchSize);<a name="line.745"></a>
-<span class="sourceLineNo">746</span>  }<a name="line.746"></a>
-<span class="sourceLineNo">747</span><a name="line.747"></a>
-<span class="sourceLineNo">748</span>  /**<a name="line.748"></a>
-<span class="sourceLineNo">749</span>   * Add HBase and its dependencies (only) to the job configuration.<a name="line.749"></a>
-<span class="sourceLineNo">750</span>   * &lt;p&gt;<a name="line.750"></a>
-<span class="sourceLineNo">751</span>   * This is intended as a low-level API, facilitating code reuse between this<a name="line.751"></a>
-<span class="sourceLineNo">752</span>   * class and its mapred counterpart. It also of use to extenral tools that<a name="line.752"></a>
-<span class="sourceLineNo">753</span>   * need to build a MapReduce job that interacts with HBase but want<a name="line.753"></a>
-<span class="sourceLineNo">754</span>   * fine-grained control over the jars shipped to the cluster.<a name="line.754"></a>
-<span class="sourceLineNo">755</span>   * &lt;/p&gt;<a name="line.755"></a>
-<span class="sourceLineNo">756</span>   * @param conf The Configuration object to extend with dependencies.<a name="line.756"></a>
-<span class="sourceLineNo">757</span>   * @see org.apache.hadoop.hbase.mapred.TableMapReduceUtil<a name="line.757"></a>
-<span class="sourceLineNo">758</span>   * @see &lt;a href="https://issues.apache.org/jira/browse/PIG-3285"&gt;PIG-3285&lt;/a&gt;<a name="line.758"></a>
-<span class="sourceLineNo">759</span>   */<a name="line.759"></a>
-<span class="sourceLineNo">760</span>  public static void addHBaseDependencyJars(Configuration conf) throws IOException {<a name="line.760"></a>
-<span class="sourceLineNo">761</span>    addDependencyJars(conf,<a name="line.761"></a>
-<span class="sourceLineNo">762</span>      // explicitly pull a class from each module<a name="line.762"></a>
-<span class="sourceLineNo">763</span>      org.apache.hadoop.hbase.HConstants.class,                      // hbase-common<a name="line.763"></a>
-<span class="sourceLineNo">764</span>      org.apache.hadoop.hbase.protobuf.generated.ClientProtos.class, // hbase-protocol<a name="line.764"></a>
-<span class="sourceLineNo">765</span>      org.apache.hadoop.hbase.client.Put.class,                      // hbase-client<a name="line.765"></a>
-<span class="sourceLineNo">766</span>      org.apache.hadoop.hbase.CompatibilityFactory.class,            // hbase-hadoop-compat<a name="line.766"></a>
-<span class="sourceLineNo">767</span>      org.apache.hadoop.hbase.mapreduce.TableMapper.class,           // hbase-server<a name="line.767"></a>
-<span class="sourceLineNo">768</span>      // pull necessary dependencies<a name="line.768"></a>
-<span class="sourceLineNo">769</span>      org.apache.zookeeper.ZooKeeper.class,<a name="line.769"></a>
-<span class="sourceLineNo">770</span>      io.netty.channel.Channel.class,<a name="line.770"></a>
-<span class="sourceLineNo">771</span>      com.google.protobuf.Message.class,<a name="line.771"></a>
-<span class="sourceLineNo">772</span>      com.google.common.collect.Lists.class,<a name="line.772"></a>
-<span class="sourceLineNo">773</span>      org.apache.htrace.Trace.class,<a name="line.773"></a>
-<span class="sourceLineNo">774</span>      com.yammer.metrics.core.MetricsRegistry.class);<a name="line.774"></a>
-<span class="sourceLineNo">775</span>  }<a name="line.775"></a>
-<span class="sourceLineNo">776</span><a name="line.776"></a>
-<span class="sourceLineNo">777</span>  /**<a name="line.777"></a>
-<span class="sourceLineNo">778</span>   * Returns a classpath string built from the content of the "tmpjars" value in {@code conf}.<a name="line.778"></a>
-<span class="sourceLineNo">779</span>   * Also exposed to shell scripts via `bin/hbase mapredcp`.<a name="line.779"></a>
-<span class="sourceLineNo">780</span>   */<a name="line.780"></a>
-<span class="sourceLineNo">781</span>  public static String buildDependencyClasspath(Configuration conf) {<a name="line.781"></a>
-<span class="sourceLineNo">782</span>    if (conf == null) {<a name="line.782"></a>
-<span class="sourceLineNo">783</span>      throw new IllegalArgumentException("Must provide a configuration object.");<a name="line.783"></a>
-<span class="sourceLineNo">784</span>    }<a name="line.784"></a>
-<span class="sourceLineNo">785</span>    Set&lt;String&gt; paths = new HashSet&lt;String&gt;(conf.getStringCollection("tmpjars"));<a name="line.785"></a>
-<span class="sourceLineNo">786</span>    if (paths.size() == 0) {<a name="line.786"></a>
-<span class="sourceLineNo">787</span>      throw new IllegalArgumentException("Configuration contains no tmpjars.");<a name="line.787"></a>
+<span class="sourceLineNo">702</span>    if (addDependencyJars) {<a name="line.702"></a>
+<span class="sourceLineNo">703</span>      addDependencyJars(job);<a name="line.703"></a>
+<span class="sourceLineNo">704</span>    }<a name="line.704"></a>
+<span class="sourceLineNo">705</span><a name="line.705"></a>
+<span class="sourceLineNo">706</span>    initCredentials(job);<a name="line.706"></a>
+<span class="sourceLineNo">707</span>  }<a name="line.707"></a>
+<span class="sourceLineNo">708</span><a name="line.708"></a>
+<span class="sourceLineNo">709</span>  /**<a name="line.709"></a>
+<span class="sourceLineNo">710</span>   * Ensures that the given number of reduce tasks for the given job<a name="line.710"></a>
+<span class="sourceLineNo">711</span>   * configuration does not exceed the number of regions for the given table.<a name="line.711"></a>
+<span class="sourceLineNo">712</span>   *<a name="line.712"></a>
+<span class="sourceLineNo">713</span>   * @param table  The table to get the region count for.<a name="line.713"></a>
+<span class="sourceLineNo">714</span>   * @param job  The current job to adjust.<a name="line.714"></a>
+<span class="sourceLineNo">715</span>   * @throws IOException When retrieving the table details fails.<a name="line.715"></a>
+<span class="sourceLineNo">716</span>   */<a name="line.716"></a>
+<span class="sourceLineNo">717</span>  public static void limitNumReduceTasks(String table, Job job)<a name="line.717"></a>
+<span class="sourceLineNo">718</span>  throws IOException {<a name="line.718"></a>
+<span class="sourceLineNo">719</span>    int regions =<a name="line.719"></a>
+<span class="sourceLineNo">720</span>      MetaTableAccessor.getRegionCount(job.getConfiguration(), TableName.valueOf(table));<a name="line.720"></a>
+<span class="sourceLineNo">721</span>    if (job.getNumReduceTasks() &gt; regions)<a name="line.721"></a>
+<span class="sourceLineNo">722</span>      job.setNumReduceTasks(regions);<a name="line.722"></a>
+<span class="sourceLineNo">723</span>  }<a name="line.723"></a>
+<span class="sourceLineNo">724</span><a name="line.724"></a>
+<span class="sourceLineNo">725</span>  /**<a name="line.725"></a>
+<span class="sourceLineNo">726</span>   * Sets the number of reduce tasks for the given job configuration to the<a name="line.726"></a>
+<span class="sourceLineNo">727</span>   * number of regions the given table has.<a name="line.727"></a>
+<span class="sourceLineNo">728</span>   *<a name="line.728"></a>
+<span class="sourceLineNo">729</span>   * @param table  The table to get the region count for.<a name="line.729"></a>
+<span class="sourceLineNo">730</span>   * @param job  The current job to adjust.<a name="line.730"></a>
+<span class="sourceLineNo">731</span>   * @throws IOException When retrieving the table details fails.<a name="line.731"></a>
+<span class="sourceLineNo">732</span>   */<a name="line.732"></a>
+<span class="sourceLineNo">733</span>  public static void setNumReduceTasks(String table, Job job)<a name="line.733"></a>
+<span class="sourceLineNo">734</span>  throws IOException {<a name="line.734"></a>
+<span class="sourceLineNo">735</span>    job.setNumReduceTasks(MetaTableAccessor.getRegionCount(job.getConfiguration(),<a name="line.735"></a>
+<span class="sourceLineNo">736</span>       TableName.valueOf(table)));<a name="line.736"></a>
+<span class="sourceLineNo">737</span>  }<a name="line.737"></a>
+<span class="sourceLineNo">738</span><a name="line.738"></a>
+<span class="sourceLineNo">739</span>  /**<a name="line.739"></a>
+<span class="sourceLineNo">740</span>   * Sets the number of rows to return and cache with each scanner iteration.<a name="line.740"></a>
+<span class="sourceLineNo">741</span>   * Higher caching values will enable faster mapreduce jobs at the expense of<a name="line.741"></a>
+<span class="sourceLineNo">742</span>   * requiring more heap to contain the cached rows.<a name="line.742"></a>
+<span class="sourceLineNo">743</span>   *<a name="line.743"></a>
+<span class="sourceLineNo">744</span>   * @param job The current job to adjust.<a name="line.744"></a>
+<span class="sourceLineNo">745</span>   * @param batchSize The number of rows to return in batch with each scanner<a name="line.745"></a>
+<span class="sourceLineNo">746</span>   * iteration.<a name="line.746"></a>
+<span class="sourceLineNo">747</span>   */<a name="line.747"></a>
+<span class="sourceLineNo">748</span>  public static void setScannerCaching(Job job, int batchSize) {<a name="line.748"></a>
+<span class="sourceLineNo">749</span>    job.getConfiguration().setInt("hbase.client.scanner.caching", batchSize);<a name="line.749"></a>
+<span class="sourceLineNo">750</span>  }<a name="line.750"></a>
+<span class="sourceLineNo">751</span><a name="line.751"></a>
+<span class="sourceLineNo">752</span>  /**<a name="line.752"></a>
+<span class="sourceLineNo">753</span>   * Add HBase and its dependencies (only) to the job configuration.<a name="line.753"></a>
+<span class="sourceLineNo">754</span>   * &lt;p&gt;<a name="line.754"></a>
+<span class="sourceLineNo">755</span>   * This is intended as a low-level API, facilitating code reuse between this<a name="line.755"></a>
+<span class="sourceLineNo">756</span>   * class and its mapred counterpart. It also of use to extenral tools that<a name="line.756"></a>
+<span class="sourceLineNo">757</span>   * need to build a MapReduce job that interacts with HBase but want<a name="line.757"></a>
+<span class="sourceLineNo">758</span>   * fine-grained control over the jars shipped to the cluster.<a name="line.758"></a>
+<span class="sourceLineNo">759</span>   * &lt;/p&gt;<a name="line.759"></a>
+<span class="sourceLineNo">760</span>   * @param conf The Configuration object to extend with dependencies.<a name="line.760"></a>
+<span class="sourceLineNo">761</span>   * @see org.apache.hadoop.hbase.mapred.TableMapReduceUtil<a name="line.761"></a>
+<span class="sourceLineNo">762</span>   * @see &lt;a href="https://issues.apache.org/jira/browse/PIG-3285"&gt;PIG-3285&lt;/a&gt;<a name="line.762"></a>
+<span class="sourceLineNo">763</span>   */<a name="line.763"></a>
+<span class="sourceLineNo">764</span>  public static void addHBaseDependencyJars(Configuration conf) throws IOException {<a name="line.764"></a>
+<span class="sourceLineNo">765</span>    addDependencyJars(conf,<a name="line.765"></a>
+<span class="sourceLineNo">766</span>      // explicitly pull a class from each module<a name="line.766"></a>
+<span class="sourceLineNo">767</span>      org.apache.hadoop.hbase.HConstants.class,                      // hbase-common<a name="line.767"></a>
+<span class="sourceLineNo">768</span>      org.apache.hadoop.hbase.protobuf.generated.ClientProtos.class, // hbase-protocol<a name="line.768"></a>
+<span class="sourceLineNo">769</span>      org.apache.hadoop.hbase.client.Put.class,                      // hbase-client<a name="line.769"></a>
+<span class="sourceLineNo">770</span>      org.apache.hadoop.hbase.CompatibilityFactory.class,            // hbase-hadoop-compat<a name="line.770"></a>
+<span class="sourceLineNo">771</span>      org.apache.hadoop.hbase.mapreduce.TableMapper.class,           // hbase-server<a name="line.771"></a>
+<span class="sourceLineNo">772</span>      // pull necessary dependencies<a name="line.772"></a>
+<span class="sourceLineNo">773</span>      org.apache.zookeeper.ZooKeeper.class,<a name="line.773"></a>
+<span class="sourceLineNo">774</span>      io.netty.channel.Channel.class,<a name="line.774"></a>
+<span class="sourceLineNo">775</span>      com.google.protobuf.Message.class,<a name="line.775"></a>
+<span class="sourceLineNo">776</span>      com.google.common.collect.Lists.class,<a name="line.776"></a>
+<span class="sourceLineNo">777</span>      org.apache.htrace.Trace.class,<a name="line.777"></a>
+<span class="sourceLineNo">778</span>      com.yammer.metrics.core.MetricsRegistry.class);<a name="line.778"></a>
+<span class="sourceLineNo">779</span>  }<a name="line.779"></a>
+<span class="sourceLineNo">780</span><a name="line.780"></a>
+<span class="sourceLineNo">781</span>  /**<a name="line.781"></a>
+<span class="sourceLineNo">782</span>   * Returns a classpath string built from the content of the "tmpjars" value in {@code conf}.<a name="line.782"></a>
+<span class="sourceLineNo">783</span>   * Also exposed to shell scripts via `bin/hbase mapredcp`.<a name="line.783"></a>
+<span class="sourceLineNo">784</span>   */<a name="line.784"></a>
+<span class="sourceLineNo">785</span>  public static String buildDependencyClasspath(Configuration conf) {<a name="line.785"></a>
+<span class="sourceLineNo">786</span>    if (conf == null) {<a name="line.786"></a>
+<span class="sourceLineNo">787</span>      throw new IllegalArgumentException("Must provide a configuration object.");<a name="line.787"></a>
 <span class="sourceLineNo">788</span>    }<a name="line.788"></a>
-<span class="sourceLineNo">789</span>    StringBuilder sb = new StringBuilder();<a name="line.789"></a>
-<span class="sourceLineNo">790</span>    for (String s : paths) {<a name="line.790"></a>
-<span class="sourceLineNo">791</span>      // entries can take the form 'file:/path/to/file.jar'.<a name="line.791"></a>
-<span class="sourceLineNo">792</span>      int idx = s.indexOf(":");<a name="line.792"></a>
-<span class="sourceLineNo">793</span>      if (idx != -1) s = s.substring(idx + 1);<a name="line.793"></a>
-<span class="sourceLineNo">794</span>      if (sb.length() &gt; 0) sb.append(File.pathSeparator);<a name="line.794"></a>
-<span class="sourceLineNo">795</span>      sb.append(s);<a name="line.795"></a>
-<span class="sourceLineNo">796</span>    }<a name="line.796"></a>
-<span class="sourceLineNo">797</span>    return sb.toString();<a name="line.797"></a>
-<span class="sourceLineNo">798</span>  }<a name="line.798"></a>
-<span class="sourceLineNo">799</span><a name="line.799"></a>
-<span class="sourceLineNo">800</span>  /**<a name="line.800"></a>
-<span class="sourceLineNo">801</span>   * Add the HBase dependency jars as well as jars for any of the configured<a name="line.801"></a>
-<span class="sourceLineNo">802</span>   * job classes to the job configuration, so that JobClient will ship them<a name="line.802"></a>
-<span class="sourceLineNo">803</span>   * to the cluster and add them to the DistributedCache.<a name="line.803"></a>
-<span class="sourceLineNo">804</span>   */<a name="line.804"></a>
-<span class="sourceLineNo">805</span>  public static void addDependencyJars(Job job) throws IOException {<a name="line.805"></a>
-<span class="sourceLineNo">806</span>    addHBaseDependencyJars(job.getConfiguration());<a name="line.806"></a>
-<span class="sourceLineNo">807</span>    try {<a name="line.807"></a>
-<span class="sourceLineNo">808</span>      addDependencyJars(job.getConfiguration(),<a name="line.808"></a>
-<span class="sourceLineNo">809</span>          // when making changes here, consider also mapred.TableMapReduceUtil<a name="line.809"></a>
-<span class="sourceLineNo">810</span>          // pull job classes<a name="line.810"></a>
-<span class="sourceLineNo">811</span>          job.getMapOutputKeyClass(),<a name="line.811"></a>
-<span class="sourceLineNo">812</span>          job.getMapOutputValueClass(),<a name="line.812"></a>
-<span class="sourceLineNo">813</span>          job.getInputFormatClass(),<a name="line.813"></a>
-<span class="sourceLineNo">814</span>          job.getOutputKeyClass(),<a name="line.814"></a>
-<span class="sourceLineNo">815</span>          job.getOutputValueClass(),<a name="line.815"></a>
-<span class="sourceLineNo">816</span>          job.getOutputFormatClass(),<a name="line.816"></a>
-<span class="sourceLineNo">817</span>          job.getPartitionerClass(),<a name="line.817"></a>
-<span class="sourceLineNo">818</span>          job.getCombinerClass());<a name="line.818"></a>
-<span class="sourceLineNo">819</span>    } catch (ClassNotFoundException e) {<a name="line.819"></a>
-<span class="sourceLineNo">820</span>      throw new IOException(e);<a name="line.820"></a>
-<span class="sourceLineNo">821</span>    }<a name="line.821"></a>
-<span class="sourceLineNo">822</span>  }<a name="line.822"></a>
-<span class="sourceLineNo">823</span><a name="line.823"></a>
-<span class="sourceLineNo">824</span>  /**<a name="line.824"></a>
-<span class="sourceLineNo">825</span>   * Add the jars containing the given classes to the job's configuration<a name="line.825"></a>
-<span class="sourceLineNo">826</span>   * such that JobClient will ship them to the cluster and add them to<a name="line.826"></a>
-<span class="sourceLineNo">827</span>   * the DistributedCache.<a name="line.827"></a>
-<span class="sourceLineNo">828</span>   */<a name="line.828"></a>
-<span class="sourceLineNo">829</span>  public static void addDependencyJars(Configuration conf,<a name="line.829"></a>
-<span class="sourceLineNo">830</span>      Class&lt;?&gt;... classes) throws IOException {<a name="line.830"></a>
-<span class="sourceLineNo">831</span><a name="line.831"></a>
-<span class="sourceLineNo">832</span>    FileSystem localFs = FileSystem.getLocal(conf);<a name="line.832"></a>
-<span class="sourceLineNo">833</span>    Set&lt;String&gt; jars = new HashSet&lt;String&gt;();<a name="line.833"></a>
-<span class="sourceLineNo">834</span>    // Add jars that are already in the tmpjars variable<a name="line.834"></a>
-<span class="sourceLineNo">835</span>    jars.addAll(conf.getStringCollection("tmpjars"));<a name="line.835"></a>
-<span class="sourceLineNo">836</span><a name="line.836"></a>
-<span class="sourceLineNo">837</span>    // add jars as we find them to a map of contents jar name so that we can avoid<a name="line.837"></a>
-<span class="sourceLineNo">838</span>    // creating new jars for classes that have already been packaged.<a name="line.838"></a>
-<span class="sourceLineNo">839</span>    Map&lt;String, String&gt; packagedClasses = new HashMap&lt;String, String&gt;();<a name="line.839"></a>
+<span class="sourceLineNo">789</span>    Set&lt;String&gt; paths = new HashSet&lt;String&gt;(conf.getStringCollection("tmpjars"));<a name="line.789"></a>
+<span class="sourceLineNo">790</span>    if (paths.size() == 0) {<a name="line.790"></a>
+<span class="sourceLineNo">791</span>      throw new IllegalArgumentException("Configuration contains no tmpjars.");<a name="line.791"></a>
+<span class="sourceLineNo">792</span>    }<a name="line.792"></a>
+<span class="sourceLineNo">793</span>    StringBuilder sb = new StringBuilder();<a name="line.793"></a>
+<span class="sourceLineNo">794</span>    for (String s : paths) {<a name="line.794"></a>
+<span class="sourceLineNo">795</span>      // entries can take the form 'file:/path/to/file.jar'.<a name="line.795"></a>
+<span class="sourceLineNo">796</span>      int idx = s.indexOf(":");<a name="line.796"></a>
+<span class="sourceLineNo">797</span>      if (idx != -1) s = s.substring(idx + 1);<a name="line.797"></a>
+<span class="sourceLineNo">798</span>      if (sb.length() &gt; 0) sb.append(File.pathSeparator);<a name="line.798"></a>
+<span class="sourceLineNo">799</span>      sb.append(s);<a name="line.799"></a>
+<span class="sourceLineNo">800</span>    }<a name="line.800"></a>
+<span class="sourceLineNo">801</span>    return sb.toString();<a name="line.801"></a>
+<span class="sourceLineNo">802</span>  }<a name="line.802"></a>
+<span class="sourceLineNo">803</span><a name="line.803"></a>
+<span class="sourceLineNo">804</span>  /**<a name="line.804"></a>
+<span class="sourceLineNo">805</span>   * Add the HBase dependency jars as well as jars for any of the configured<a name="line.805"></a>
+<span class="sourceLineNo">806</span>   * job classes to the job configuration, so that JobClient will ship them<a name="line.806"></a>
+<span class="sourceLineNo">807</span>   * to the cluster and add them to the DistributedCache.<a name="line.807"></a>
+<span class="sourceLineNo">808</span>   */<a name="line.808"></a>
+<span class="sourceLineNo">809</span>  public static void addDependencyJars(Job job) throws IOException {<a name="line.809"></a>
+<span class="sourceLineNo">810</span>    addHBaseDependencyJars(job.getConfiguration());<a name="line.810"></a>
+<span class="sourceLineNo">811</span>    try {<a name="line.811"></a>
+<span class="sourceLineNo">812</span>      addDependencyJars(job.getConfiguration(),<a name="line.812"></a>
+<span class="sourceLineNo">813</span>          // when making changes here, consider also mapred.TableMapReduceUtil<a name="line.813"></a>
+<span class="sourceLineNo">814</span>          // pull job classes<a name="line.814"></a>
+<span class="sourceLineNo">815</span>          job.getMapOutputKeyClass(),<a name="line.815"></a>
+<span class="sourceLineNo">816</span>          job.getMapOutputValueClass(),<a name="line.816"></a>
+<span class="sourceLineNo">817</span>          job.getInputFormatClass(),<a name="line.817"></a>
+<span class="sourceLineNo">818</span>          job.getOutputKeyClass(),<a name="line.818"></a>
+<span class="sourceLineNo">819</span>          job.getOutputValueClass(),<a name="line.819"></a>
+<span class="sourceLineNo">820</span>          job.getOutputFormatClass(),<a name="line.820"></a>
+<span class="sourceLineNo">821</span>          job.getPartitionerClass(),<a name="line.821"></a>
+<span class="sourceLineNo">822</span>          job.getCombinerClass());<a name="line.822"></a>
+<span class="sourceLineNo">823</span>    } catch (ClassNotFoundException e) {<a name="line.823"></a>
+<span class="sourceLineNo">824</span>      throw new IOException(e);<a name="line.824"></a>
+<span class="sourceLineNo">825</span>    }<a name="line.825"></a>
+<span class="sourceLineNo">826</span>  }<a name="line.826"></a>
+<span class="sourceLineNo">827</span><a name="line.827"></a>
+<span class="sourceLineNo">828</span>  /**<a name="line.828"></a>
+<span class="sourceLineNo">829</span>   * Add the jars containing the given classes to the job's configuration<a name="line.829"></a>
+<span class="sourceLineNo">830</span>   * such that JobClient will ship them to the cluster and add them to<a name="line.830"></a>
+<span class="sourceLineNo">831</span>   * the DistributedCache.<a name="line.831"></a>
+<span class="sourceLineNo">832</span>   */<a name="line.832"></a>
+<span class="sourceLineNo">833</span>  public static void addDependencyJars(Configuration conf,<a name="line.833"></a>
+<span class="sourceLineNo">834</span>      Class&lt;?&gt;... classes) throws IOException {<a name="line.834"></a>
+<span class="sourceLineNo">835</span><a name="line.835"></a>
+<span class="sourceLineNo">836</span>    FileSystem localFs = FileSystem.getLocal(conf);<a name="line.836"></a>
+<span class="sourceLineNo">837</span>    Set&lt;String&gt; jars = new HashSet&lt;String&gt;();<a name="line.837"></a>
+<span class="sourceLineNo">838</span>    // Add jars that are already in the tmpjars variable<a name="line.838"></a>
+<span class="sourceLineNo">839</span>    jars.addAll(conf.getStringCollection("tmpjars"));<a name="line.839"></a>
 <span class="sourceLineNo">840</span><a name="line.840"></a>
-<span class="sourceLineNo">841</span>    // Add jars containing the specified classes<a name="line.841"></a>
-<span class="sourceLineNo">842</span>    for (Class&lt;?&gt; clazz : classes) {<a name="line.842"></a>
-<span class="sourceLineNo">843</span>      if (clazz == null) continue;<a name="line.843"></a>
+<span class="sourceLineNo">841</span>    // add jars as we find them to a map of contents jar name so that we can avoid<a name="line.841"></a>
+<span class="sourceLineNo">842</span>    // creating new jars for classes that have already been packaged.<a name="line.842"></a>
+<span class="sourceLineNo">843</span>    Map&lt;String, String&gt; packagedClasses = new HashMap&lt;String, String&gt;();<a name="line.843"></a>
 <span class="sourceLineNo">844</span><a name="line.844"></a>
-<span class="sourceLineNo">845</span>      Path path = findOrCreateJar(clazz, localFs, packagedClasses);<a name="line.845"></a>
-<span class="sourceLineNo">846</span>      if (path == null) {<a name="line.846"></a>
-<span class="sourceLineNo">847</span>        LOG.warn("Could not find jar for class " + clazz +<a name="line.847"></a>
-<span class="sourceLineNo">848</span>                 " in order to ship it to the cluster.");<a name="line.848"></a>
-<span class="sourceLineNo">849</span>        continue;<a name="line.849"></a>
-<span class="sourceLineNo">850</span>      }<a name="line.850"></a>
-<span class="sourceLineNo">851</span>      if (!localFs.exists(path)) {<a name="line.851"></a>
-<span class="sourceLineNo">852</span>        LOG.warn("Could not validate jar file " + path + " for class "<a name="line.852"></a>
-<span class="sourceLineNo">853</span>                 + clazz);<a name="line.853"></a>
-<span class="sourceLineNo">854</span>        continue;<a name="line.854"></a>
-<span class="sourceLineNo">855</span>      }<a name="line.855"></a>
-<span class="sourceLineNo">856</span>      jars.add(path.toString());<a name="line.856"></a>
-<span class="sourceLineNo">857</span>    }<a name="line.857"></a>
-<span class="sourceLineNo">858</span>    if (jars.isEmpty()) return;<a name="line.858"></a>
-<span class="sourceLineNo">859</span><a name="line.859"></a>
-<span class="sourceLineNo">860</span>    conf.set("tmpjars", StringUtils.arrayToString(jars.toArray(new String[jars.size()])));<a name="line.860"></a>
-<span class="sourceLineNo">861</span>  }<a name="line.861"></a>
-<span class="sourceLineNo">862</span><a name="line.862"></a>
-<span class="sourceLineNo">863</span>  /**<a name="line.863"></a>
-<span class="sourceLineNo">864</span>   * Finds the Jar for a class or creates it if it doesn't exist. If the class is in<a name="line.864"></a>
-<span class="sourceLineNo">865</span>   * a directory in the classpath, it creates a Jar on the fly with the<a name="line.865"></a>
-<span class="sourceLineNo">866</span>   * contents of the directory and returns the path to that Jar. If a Jar is<a name="line.866"></a>
-<span class="sourceLineNo">867</span>   * created, it is created in the system temporary directory. Otherwise,<a name="line.867"></a>
-<span class="sourceLineNo">868</span>   * returns an existing jar that contains a class of the same name. Maintains<a name="line.868"></a>
-<span class="sourceLineNo">869</span>   * a mapping from jar contents to the tmp jar created.<a name="line.869"></a>
-<span class="sourceLineNo">870</span>   * @param my_class the class to find.<a name="line.870"></a>
-<span class="sourceLineNo">871</span>   * @param fs the FileSystem with which to qualify the returned path.<a name="line.871"></a>
-<span class="sourceLineNo">872</span>   * @param packagedClasses a map of class name to path.<a name="line.872"></a>
-<span class="sourceLineNo">873</span>   * @return a jar file that contains the class.<a name="line.873"></a>
-<span class="sourceLineNo">874</span>   * @throws IOException<a name="line.874"></a>
-<span class="sourceLineNo">875</span>   */<a name="line.875"></a>
-<span class="sourceLineNo">876</span>  private static Path findOrCreateJar(Class&lt;?&gt; my_class, FileSystem fs,<a name="line.876"></a>
-<span class="sourceLineNo">877</span>      Map&lt;String, String&gt; packagedClasses)<a name="line.877"></a>
-<span class="sourceLineNo">878</span>  throws IOException {<a name="line.878"></a>
-<span class="sourceLineNo">879</span>    // attempt to locate an existing jar for the class.<a name="line.879"></a>
-<span class="sourceLineNo">880</span>    String jar = findContainingJar(my_class, packagedClasses);<a name="line.880"></a>
-<span class="sourceLineNo">881</span>    if (null == jar || jar.isEmpty()) {<a name="line.881"></a>
-<span class="sourceLineNo">882</span>      jar = getJar(my_class);<a name="line.882"></a>
-<span class="sourceLineNo">883</span>      updateMap(jar, packagedClasses);<a name="line.883"></a>
-<span class="sourceLineNo">884</span>    }<a name="line.884"></a>
-<span class="sourceLineNo">885</span><a name="line.885"></a>
-<span class="sourceLineNo">886</span>    if (null == jar || jar.isEmpty()) {<a name="line.886"></a>
-<span class="sourceLineNo">887</span>      return null;<a name="line.887"></a>
+<span class="sourceLineNo">845</span>    // Add jars containing the specified classes<a name="line.845"></a>
+<span class="sourceLineNo">846</span>    for (Class&lt;?&gt; clazz : classes) {<a name="line.846"></a>
+<span class="sourceLineNo">847</span>      if (clazz == null) continue;<a name="line.847"></a>
+<span class="sourceLineNo">848</span><a name="line.848"></a>
+<span class="sourceLineNo">849</span>      Path path = findOrCreateJar(clazz, localFs, packagedClasses);<a name="line.849"></a>
+<span class="sourceLineNo">850</span>      if (path == null) {<a name="line.850"></a>
+<span class="sourceLineNo">851</span>        LOG.warn("Could not find jar for class " + clazz +<a name="line.851"></a>
+<span class="sourceLineNo">852</span>                 " in order to ship it to the cluster.");<a name="line.852"></a>
+<span class="sourceLineNo">853</span>        continue;<a name="line.853"></a>
+<span class="sourceLineNo">854</span>      }<a name="line.854"></a>
+<span class="sourceLineNo">855</span>      if (!localFs.exists(path)) {<a name="line.855"></a>
+<span class="sourceLineNo">856</span>        LOG.warn("Could not validate jar file " + path + " for class "<a name="line.856"></a>
+<span class="sourceLineNo">857</span>                 + clazz);<a name="line.857"></a>
+<span class="sourceLineNo">858</span>        continue;<a name="line.858"></a>
+<span class="sourceLineNo">859</span>      }<a name="line.859"></a>
+<span class="sourceLineNo">860</span>      jars.add(path.toString());<a name="line.860"></a>
+<span class="sourceLineNo">861</span>    }<a name="line.861"></a>
+<span class="sourceLineNo">862</span>    if (jars.isEmpty()) return;<a name="line.862"></a>
+<span class="sourceLineNo">863</span><a name="line.863"></a>
+<span class="sourceLineNo">864</span>    conf.set("tmpjars", StringUtils.arrayToString(jars.toArray(new String[jars.size()])));<a name="line.864"></a>
+<span class="sourceLineNo">865</span>  }<a name="line.865"></a>
+<span class="sourceLineNo">866</span><a name="line.866"></a>
+<span class="sourceLineNo">867</span>  /**<a name="line.867"></a>
+<span class="sourceLineNo">868</span>   * Finds the Jar for a class or creates it if it doesn't exist. If the class is in<a name="line.868"></a>
+<span class="sourceLineNo">869</span>   * a directory in the classpath, it creates a Jar on the fly with the<a name="line.869"></a>
+<span class="sourceLineNo">870</span>   * contents of the directory and returns the path to that Jar. If a Jar is<a name="line.870"></a>
+<span class="sourceLineNo">871</span>   * created, it is created in the system temporary directory. Otherwise,<a name="line.871"></a>
+<span class="sourceLineNo">872</span>   * returns an existing jar that contains a class of the same name. Maintains<a name="line.872"></a>
+<span class="sourceLineNo">873</span>   * a mapping from jar contents to the tmp jar created.<a name="line.873"></a>
+<span class="sourceLineNo">874</span>   * @param my_class the class to find.<a name="line.874"></a>
+<span class="sourceLineNo">875</span>   * @param fs the FileSystem with which to qualify the returned path.<a name="line.875"></a>
+<span class="sourceLineNo">876</span>   * @param packagedClasses a map of class name to path.<a name="line.876"></a>
+<span class="sourceLineNo">877</span>   * @return a jar file that contains the class.<a name="line.877"></a>
+<span class="sourceLineNo">878</span>   * @throws IOException<a name="line.878"></a>
+<span class="sourceLineNo">879</span>   */<a name="line.879"></a>
+<span class="sourceLineNo">880</span>  private static Path findOrCreateJar(Class&lt;?&gt; my_class, FileSystem fs,<a name="line.880"></a>
+<span class="sourceLineNo">881</span>      Map&lt;String, String&gt; packagedClasses)<a name="line.881"></a>
+<span class="sourceLineNo">882</span>  throws IOException {<a name="line.882"></a>
+<span class="sourceLineNo">883</span>    // attempt to locate an existing jar for the class.<a name="line.883"></a>
+<span class="sourceLineNo">884</span>    String jar = findContainingJar(my_class, packagedClasses);<a name="line.884"></a>
+<span class="sourceLineNo">885</span>    if (null == jar || jar.isEmpty()) {<a name="line.885"></a>
+<span class="sourceLineNo">886</span>      jar = getJar(my_class);<a name="line.886"></a>
+<span class="sourceLineNo">887</span>      updateMap(jar, packagedClasses);<a name="line.887"></a>
 <span class="sourceLineNo">888</span>    }<a name="line.888"></a>
 <span class="sourceLineNo">889</span><a name="line.889"></a>
-<span class="sourceLineNo">890</span>    LOG.debug(String.format("For class %s, using jar %s", my_class.getName(), jar));<a name="line.890"></a>
-<span class="sourceLineNo">891</span>    return new Path(jar).makeQualified(fs);<a name="line.891"></a>
-<span class="sourceLineNo">892</span>  }<a name="line.892"></a>
+<span class="sourceLineNo">890</span>    if (null == jar || jar.isEmpty()) {<a name="line.890"></a>
+<span class="sourceLineNo">891</span>      return null;<a name="line.891"></a>
+<span class="sourceLineNo">892</span>    }<a name="line.892"></a>
 <span class="sourceLineNo">893</span><a name="line.893"></a>
-<span class="sourceLineNo">894</span>  /**<a name="line.894"></a>
-<span class="sourceLineNo">895</span>   * Add entries to &lt;code&gt;packagedClasses&lt;/code&gt; corresponding to class files<a name="line.895"></a>
-<span class="sourceLineNo">896</span>   * contained in &lt;code&gt;jar&lt;/code&gt;.<a name="line.896"></a>
-<span class="sourceLineNo">897</span>   * @param jar The jar who's content to list.<a name="line.897"></a>
-<span class="sourceLineNo">898</span>   * @param packagedClasses map[class -&gt; jar]<a name="line.898"></a>
-<span class="sourceLineNo">899</span>   */<a name="line.899"></a>
-<span class="sourceLineNo">900</span>  private static void updateMap(String jar, Map&lt;String, String&gt; packagedClasses) throws IOException {<a name="line.900"></a>
-<span class="sourceLineNo">901</span>    if (null == jar || jar.isEmpty()) {<a name="line.901"></a>
-<span class="sourceLineNo">902</span>      return;<a name="line.902"></a>
-<span class="sourceLineNo">903</span>    }<a name="line.903"></a>
-<span class="sourceLineNo">904</span>    ZipFile zip = null;<a name="line.904"></a>
-<span class="sourceLineNo">905</span>    try {<a name="line.905"></a>
-<span class="sourceLineNo">906</span>      zip = new ZipFile(jar);<a name="line.906"></a>
-<span class="sourceLineNo">907</span>      for (Enumeration&lt;? extends ZipEntry&gt; iter = zip.entries(); iter.hasMoreElements();) {<a name="line.907"></a>
-<span class="sourceLineNo">908</span>        ZipEntry entry = iter.nextElement();<a name="line.908"></a>
-<span class="sourceLineNo">909</span>        if (entry.getName().endsWith("class")) {<a name="line.909"></a>
-<span class="sourceLineNo">910</span>          packagedClasses.put(entry.getName(), jar);<a name="line.910"></a>
-<span class="sourceLineNo">911</span>        }<a name="line.911"></a>
-<span class="sourceLineNo">912</span>      }<a name="line.912"></a>
-<span class="sourceLineNo">913</span>    } finally {<a name="line.913"></a>
-<span class="sourceLineNo">914</span>      if (null != zip) zip.close();<a name="line.914"></a>
-<span class="sourceLineNo">915</span>    }<a name="line.915"></a>
-<span class="sourceLineNo">916</span>  }<a name="line.916"></a>
-<span class="sourceLineNo">917</span><a name="line.917"></a>
-<span class="sourceLineNo">918</span>  /**<a name="line.918"></a>
-<span class="sourceLineNo">919</span>   * Find a jar that contains a class of the same name, if any. It will return<a name="line.919"></a>
-<span class="sourceLineNo">920</span>   * a jar file, even if that is not the first thing on the class path that<a name="line.920"></a>
-<span class="sourceLineNo">921</span>   * has a class with the same name. Looks first on the classpath and then in<a name="line.921"></a>
-<span class="sourceLineNo">922</span>   * the &lt;code&gt;packagedClasses&lt;/code&gt; map.<a name="line.922"></a>
-<span class="sourceLineNo">923</span>   * @param my_class the class to find.<a name="line.923"></a>
-<span class="sourceLineNo">924</span>   * @return a jar file that contains the class, or null.<a name="line.924"></a>
-<span class="sourceLineNo">925</span>   * @throws IOException<a name="line.925"></a>
-<span class="sourceLineNo">926</span>   */<a name="line.926"></a>
-<span class="sourceLineNo">927</span>  private static String findContainingJar(Class&lt;?&gt; my_class, Map&lt;String, String&gt; packagedClasses)<a name="line.927"></a>
-<span class="sourceLineNo">928</span>      throws IOException {<a name="line.928"></a>
-<span class="sourceLineNo">929</span>    ClassLoader loader = my_class.getClassLoader();<a name="line.929"></a>
-<span class="sourceLineNo">930</span><a name="line.930"></a>
-<span class="sourceLineNo">931</span>    String class_file = my_class.getName().replaceAll("\\.", "/") + ".class";<a name="line.931"></a>
-<span class="sourceLineNo">932</span><a name="line.932"></a>
-<span class="sourceLineNo">933</span>    if (loader != null) {<a name="line.933"></a>
-<span class="sourceLineNo">934</span>      // first search the classpath<a name="line.934"></a>
-<span class="sourceLineNo">935</span>      for (Enumeration&lt;URL&gt; itr = loader.getResources(class_file); itr.hasMoreElements();) {<a name="line.935"></a>
-<span class="sourceLineNo">936</span>        URL url = itr.nextElement();<a name="line.936"></a>
-<span class="sourceLineNo">937</span>        if ("jar".equals(url.getProtocol())) {<a name="line.937"></a>
-<span class="sourceLineNo">938</span>          String toReturn = url.getPath();<a name="line.938"></a>
-<span class="sourceLineNo">939</span>          if (toReturn.startsWith("file:")) {<a name="line.939"></a>
-<span class="sourceLineNo">940</span>            toReturn = toReturn.substring("file:".length());<a name="line.940"></a>
-<span class="sourceLineNo">941</span>          }<a name="line.941"></a>
-<span class="sourceLineNo">942</span>          // URLDecoder is a misnamed class, since it actually decodes<a name="line.942"></a>
-<span class="sourceLineNo">943</span>          // x-www-form-urlencoded MIME type rather than actual<a name="line.943"></a>
-<span class="sourceLineNo">944</span>          // URL encoding (which the file path has). Therefore it would<a name="line.944"></a>
-<span class="sourceLineNo">945</span>          // decode +s to ' 's which is incorrect (spaces are actually<a name="line.945"></a>
-<span class="sourceLineNo">946</span>          // either unencoded or encoded as "%20"). Replace +s first, so<a name="line.946"></a>
-<span class="sourceLineNo">947</span>          // that they are kept sacred during the decoding process.<a name="line.947"></a>
-<span class="sourceLineNo">948</span>          toReturn = toReturn.replaceAll("\\+", "%2B");<a name="line.948"></a>
-<span class="sourceLineNo">949</span>          toReturn = URLDecoder.decode(toReturn, "UTF-8");<a name="line.949"></a>
-<span class="sourceLineNo">950</span>          return toReturn.replaceAll("!.*$", "");<a name="line.950"></a>
-<span class="sourceLineNo">951</span>        }<a name="line.951"></a>
-<span class="sourceLineNo">952

<TRUNCATED>

Mime
View raw message