[47/51] [partial] hbase-site git commit: Published site at .

git-site-role Sat, 30 Sep 2017 08:14:09 -0700
http://git-wip-us.apache.org/repos/asf/hbase-site/blob/d41f56fe/apidocs/src-html/org/apache/hadoop/hbase/mapred/TableMapReduceUtil.html
----------------------------------------------------------------------
diff --git 
a/apidocs/src-html/org/apache/hadoop/hbase/mapred/TableMapReduceUtil.html 
b/apidocs/src-html/org/apache/hadoop/hbase/mapred/TableMapReduceUtil.html
index 1949f0d..3e3acbe 100644
--- a/apidocs/src-html/org/apache/hadoop/hbase/mapred/TableMapReduceUtil.html
+++ b/apidocs/src-html/org/apache/hadoop/hbase/mapred/TableMapReduceUtil.html
@@ -41,345 +41,383 @@
 <span class="sourceLineNo">033</span>import 
org.apache.hadoop.hbase.security.User;<a name="line.33"></a>
 <span class="sourceLineNo">034</span>import 
org.apache.hadoop.hbase.security.UserProvider;<a name="line.34"></a>
 <span class="sourceLineNo">035</span>import 
org.apache.hadoop.hbase.security.token.TokenUtil;<a name="line.35"></a>
-<span class="sourceLineNo">036</span>import 
org.apache.hadoop.mapred.FileInputFormat;<a name="line.36"></a>
-<span class="sourceLineNo">037</span>import 
org.apache.hadoop.mapred.InputFormat;<a name="line.37"></a>
-<span class="sourceLineNo">038</span>import 
org.apache.hadoop.mapred.JobConf;<a name="line.38"></a>
-<span class="sourceLineNo">039</span>import 
org.apache.hadoop.mapred.OutputFormat;<a name="line.39"></a>
-<span class="sourceLineNo">040</span>import 
org.apache.hadoop.mapred.TextInputFormat;<a name="line.40"></a>
-<span class="sourceLineNo">041</span>import 
org.apache.hadoop.mapred.TextOutputFormat;<a name="line.41"></a>
-<span class="sourceLineNo">042</span><a name="line.42"></a>
-<span class="sourceLineNo">043</span>import java.io.IOException;<a 
name="line.43"></a>
-<span class="sourceLineNo">044</span>import java.util.Collection;<a 
name="line.44"></a>
-<span class="sourceLineNo">045</span>import java.util.Map;<a 
name="line.45"></a>
-<span class="sourceLineNo">046</span><a name="line.46"></a>
-<span class="sourceLineNo">047</span>/**<a name="line.47"></a>
-<span class="sourceLineNo">048</span> * Utility for {@link TableMap} and 
{@link TableReduce}<a name="line.48"></a>
-<span class="sourceLineNo">049</span> */<a name="line.49"></a>
-<span class="sourceLineNo">050</span>@InterfaceAudience.Public<a 
name="line.50"></a>
-<span class="sourceLineNo">051</span>@SuppressWarnings({ "rawtypes", 
"unchecked" })<a name="line.51"></a>
-<span class="sourceLineNo">052</span>public class TableMapReduceUtil {<a 
name="line.52"></a>
-<span class="sourceLineNo">053</span><a name="line.53"></a>
-<span class="sourceLineNo">054</span>  /**<a name="line.54"></a>
-<span class="sourceLineNo">055</span>   * Use this before submitting a 
TableMap job. It will<a name="line.55"></a>
-<span class="sourceLineNo">056</span>   * appropriately set up the JobConf.<a 
name="line.56"></a>
-<span class="sourceLineNo">057</span>   *<a name="line.57"></a>
-<span class="sourceLineNo">058</span>   * @param table  The table name to read 
from.<a name="line.58"></a>
-<span class="sourceLineNo">059</span>   * @param columns  The columns to 
scan.<a name="line.59"></a>
-<span class="sourceLineNo">060</span>   * @param mapper  The mapper class to 
use.<a name="line.60"></a>
-<span class="sourceLineNo">061</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.61"></a>
-<span class="sourceLineNo">062</span>   * @param outputValueClass  The class 
of the output value.<a name="line.62"></a>
-<span class="sourceLineNo">063</span>   * @param job  The current job 
configuration to adjust.<a name="line.63"></a>
-<span class="sourceLineNo">064</span>   */<a name="line.64"></a>
-<span class="sourceLineNo">065</span>  public static void 
initTableMapJob(String table, String columns,<a name="line.65"></a>
-<span class="sourceLineNo">066</span>    Class&lt;? extends TableMap&gt; 
mapper,<a name="line.66"></a>
-<span class="sourceLineNo">067</span>    Class&lt;?&gt; outputKeyClass,<a 
name="line.67"></a>
-<span class="sourceLineNo">068</span>    Class&lt;?&gt; outputValueClass, 
JobConf job) {<a name="line.68"></a>
-<span class="sourceLineNo">069</span>    initTableMapJob(table, columns, 
mapper, outputKeyClass, outputValueClass, job,<a name="line.69"></a>
-<span class="sourceLineNo">070</span>      true, TableInputFormat.class);<a 
name="line.70"></a>
-<span class="sourceLineNo">071</span>  }<a name="line.71"></a>
-<span class="sourceLineNo">072</span><a name="line.72"></a>
-<span class="sourceLineNo">073</span>  public static void 
initTableMapJob(String table, String columns,<a name="line.73"></a>
-<span class="sourceLineNo">074</span>    Class&lt;? extends TableMap&gt; 
mapper,<a name="line.74"></a>
-<span class="sourceLineNo">075</span>    Class&lt;?&gt; outputKeyClass,<a 
name="line.75"></a>
-<span class="sourceLineNo">076</span>    Class&lt;?&gt; outputValueClass, 
JobConf job, boolean addDependencyJars) {<a name="line.76"></a>
-<span class="sourceLineNo">077</span>    initTableMapJob(table, columns, 
mapper, outputKeyClass, outputValueClass, job,<a name="line.77"></a>
-<span class="sourceLineNo">078</span>      addDependencyJars, 
TableInputFormat.class);<a name="line.78"></a>
-<span class="sourceLineNo">079</span>  }<a name="line.79"></a>
-<span class="sourceLineNo">080</span><a name="line.80"></a>
-<span class="sourceLineNo">081</span>  /**<a name="line.81"></a>
-<span class="sourceLineNo">082</span>   * Use this before submitting a 
TableMap job. It will<a name="line.82"></a>
-<span class="sourceLineNo">083</span>   * appropriately set up the JobConf.<a 
name="line.83"></a>
-<span class="sourceLineNo">084</span>   *<a name="line.84"></a>
-<span class="sourceLineNo">085</span>   * @param table  The table name to read 
from.<a name="line.85"></a>
-<span class="sourceLineNo">086</span>   * @param columns  The columns to 
scan.<a name="line.86"></a>
-<span class="sourceLineNo">087</span>   * @param mapper  The mapper class to 
use.<a name="line.87"></a>
-<span class="sourceLineNo">088</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.88"></a>
-<span class="sourceLineNo">089</span>   * @param outputValueClass  The class 
of the output value.<a name="line.89"></a>
-<span class="sourceLineNo">090</span>   * @param job  The current job 
configuration to adjust.<a name="line.90"></a>
-<span class="sourceLineNo">091</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.91"></a>
-<span class="sourceLineNo">092</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.92"></a>
-<span class="sourceLineNo">093</span>   */<a name="line.93"></a>
-<span class="sourceLineNo">094</span>  public static void 
initTableMapJob(String table, String columns,<a name="line.94"></a>
-<span class="sourceLineNo">095</span>    Class&lt;? extends TableMap&gt; 
mapper,<a name="line.95"></a>
-<span class="sourceLineNo">096</span>    Class&lt;?&gt; outputKeyClass,<a 
name="line.96"></a>
-<span class="sourceLineNo">097</span>    Class&lt;?&gt; outputValueClass, 
JobConf job, boolean addDependencyJars,<a name="line.97"></a>
-<span class="sourceLineNo">098</span>    Class&lt;? extends InputFormat&gt; 
inputFormat) {<a name="line.98"></a>
-<span class="sourceLineNo">099</span><a name="line.99"></a>
-<span class="sourceLineNo">100</span>    job.setInputFormat(inputFormat);<a 
name="line.100"></a>
-<span class="sourceLineNo">101</span>    
job.setMapOutputValueClass(outputValueClass);<a name="line.101"></a>
-<span class="sourceLineNo">102</span>    
job.setMapOutputKeyClass(outputKeyClass);<a name="line.102"></a>
-<span class="sourceLineNo">103</span>    job.setMapperClass(mapper);<a 
name="line.103"></a>
-<span class="sourceLineNo">104</span>    job.setStrings("io.serializations", 
job.get("io.serializations"),<a name="line.104"></a>
-<span class="sourceLineNo">105</span>        
MutationSerialization.class.getName(), ResultSerialization.class.getName());<a 
name="line.105"></a>
-<span class="sourceLineNo">106</span>    FileInputFormat.addInputPaths(job, 
table);<a name="line.106"></a>
-<span class="sourceLineNo">107</span>    job.set(TableInputFormat.COLUMN_LIST, 
columns);<a name="line.107"></a>
-<span class="sourceLineNo">108</span>    if (addDependencyJars) {<a 
name="line.108"></a>
-<span class="sourceLineNo">109</span>      try {<a name="line.109"></a>
-<span class="sourceLineNo">110</span>        addDependencyJars(job);<a 
name="line.110"></a>
-<span class="sourceLineNo">111</span>      } catch (IOException e) {<a 
name="line.111"></a>
-<span class="sourceLineNo">112</span>        e.printStackTrace();<a 
name="line.112"></a>
-<span class="sourceLineNo">113</span>      }<a name="line.113"></a>
-<span class="sourceLineNo">114</span>    }<a name="line.114"></a>
-<span class="sourceLineNo">115</span>    try {<a name="line.115"></a>
-<span class="sourceLineNo">116</span>      initCredentials(job);<a 
name="line.116"></a>
-<span class="sourceLineNo">117</span>    } catch (IOException ioe) {<a 
name="line.117"></a>
-<span class="sourceLineNo">118</span>      // just spit out the stack trace?  
really?<a name="line.118"></a>
-<span class="sourceLineNo">119</span>      ioe.printStackTrace();<a 
name="line.119"></a>
-<span class="sourceLineNo">120</span>    }<a name="line.120"></a>
-<span class="sourceLineNo">121</span>  }<a name="line.121"></a>
-<span class="sourceLineNo">122</span><a name="line.122"></a>
-<span class="sourceLineNo">123</span>  /**<a name="line.123"></a>
-<span class="sourceLineNo">124</span>   * Sets up the job for reading from one 
or more multiple table snapshots, with one or more scans<a name="line.124"></a>
-<span class="sourceLineNo">125</span>   * per snapshot.<a name="line.125"></a>
-<span class="sourceLineNo">126</span>   * It bypasses hbase servers and read 
directly from snapshot files.<a name="line.126"></a>
-<span class="sourceLineNo">127</span>   *<a name="line.127"></a>
-<span class="sourceLineNo">128</span>   * @param snapshotScans     map of 
snapshot name to scans on that snapshot.<a name="line.128"></a>
-<span class="sourceLineNo">129</span>   * @param mapper            The mapper 
class to use.<a name="line.129"></a>
-<span class="sourceLineNo">130</span>   * @param outputKeyClass    The class 
of the output key.<a name="line.130"></a>
-<span class="sourceLineNo">131</span>   * @param outputValueClass  The class 
of the output value.<a name="line.131"></a>
-<span class="sourceLineNo">132</span>   * @param job               The current 
job to adjust.  Make sure the passed job is<a name="line.132"></a>
-<span class="sourceLineNo">133</span>   *                          carrying 
all necessary HBase configuration.<a name="line.133"></a>
-<span class="sourceLineNo">134</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.134"></a>
-<span class="sourceLineNo">135</span>   *                          job classes 
via the distributed cache (tmpjars).<a name="line.135"></a>
-<span class="sourceLineNo">136</span>   */<a name="line.136"></a>
-<span class="sourceLineNo">137</span>  public static void 
initMultiTableSnapshotMapperJob(Map&lt;String, Collection&lt;Scan&gt;&gt; 
snapshotScans,<a name="line.137"></a>
-<span class="sourceLineNo">138</span>      Class&lt;? extends TableMap&gt; 
mapper, Class&lt;?&gt; outputKeyClass, Class&lt;?&gt; outputValueClass,<a 
name="line.138"></a>
-<span class="sourceLineNo">139</span>      JobConf job, boolean 
addDependencyJars, Path tmpRestoreDir) throws IOException {<a 
name="line.139"></a>
-<span class="sourceLineNo">140</span>    
MultiTableSnapshotInputFormat.setInput(job, snapshotScans, tmpRestoreDir);<a 
name="line.140"></a>
-<span class="sourceLineNo">141</span><a name="line.141"></a>
-<span class="sourceLineNo">142</span>    
job.setInputFormat(MultiTableSnapshotInputFormat.class);<a name="line.142"></a>
-<span class="sourceLineNo">143</span>    if (outputValueClass != null) {<a 
name="line.143"></a>
-<span class="sourceLineNo">144</span>      
job.setMapOutputValueClass(outputValueClass);<a name="line.144"></a>
-<span class="sourceLineNo">145</span>    }<a name="line.145"></a>
-<span class="sourceLineNo">146</span>    if (outputKeyClass != null) {<a 
name="line.146"></a>
-<span class="sourceLineNo">147</span>      
job.setMapOutputKeyClass(outputKeyClass);<a name="line.147"></a>
-<span class="sourceLineNo">148</span>    }<a name="line.148"></a>
-<span class="sourceLineNo">149</span>    job.setMapperClass(mapper);<a 
name="line.149"></a>
-<span class="sourceLineNo">150</span>    if (addDependencyJars) {<a 
name="line.150"></a>
-<span class="sourceLineNo">151</span>      addDependencyJars(job);<a 
name="line.151"></a>
-<span class="sourceLineNo">152</span>    }<a name="line.152"></a>
-<span class="sourceLineNo">153</span><a name="line.153"></a>
-<span class="sourceLineNo">154</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.resetCacheConfig(job);<a 
name="line.154"></a>
-<span class="sourceLineNo">155</span>  }<a name="line.155"></a>
-<span class="sourceLineNo">156</span><a name="line.156"></a>
-<span class="sourceLineNo">157</span>  /**<a name="line.157"></a>
-<span class="sourceLineNo">158</span>   * Sets up the job for reading from a 
table snapshot. It bypasses hbase servers<a name="line.158"></a>
-<span class="sourceLineNo">159</span>   * and read directly from snapshot 
files.<a name="line.159"></a>
-<span class="sourceLineNo">160</span>   *<a name="line.160"></a>
-<span class="sourceLineNo">161</span>   * @param snapshotName The name of the 
snapshot (of a table) to read from.<a name="line.161"></a>
-<span class="sourceLineNo">162</span>   * @param columns  The columns to 
scan.<a name="line.162"></a>
-<span class="sourceLineNo">163</span>   * @param mapper  The mapper class to 
use.<a name="line.163"></a>
-<span class="sourceLineNo">164</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.164"></a>
-<span class="sourceLineNo">165</span>   * @param outputValueClass  The class 
of the output value.<a name="line.165"></a>
-<span class="sourceLineNo">166</span>   * @param job  The current job to 
adjust.  Make sure the passed job is<a name="line.166"></a>
-<span class="sourceLineNo">167</span>   * carrying all necessary HBase 
configuration.<a name="line.167"></a>
-<span class="sourceLineNo">168</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.168"></a>
-<span class="sourceLineNo">169</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.169"></a>
-<span class="sourceLineNo">170</span>   * @param tmpRestoreDir a temporary 
directory to copy the snapshot files into. Current user should<a 
name="line.170"></a>
-<span class="sourceLineNo">171</span>   * have write permissions to this 
directory, and this should not be a subdirectory of rootdir.<a 
name="line.171"></a>
-<span class="sourceLineNo">172</span>   * After the job is finished, restore 
directory can be deleted.<a name="line.172"></a>
-<span class="sourceLineNo">173</span>   * @throws IOException When setting up 
the details fails.<a name="line.173"></a>
-<span class="sourceLineNo">174</span>   * @see TableSnapshotInputFormat<a 
name="line.174"></a>
-<span class="sourceLineNo">175</span>   */<a name="line.175"></a>
-<span class="sourceLineNo">176</span>  public static void 
initTableSnapshotMapJob(String snapshotName, String columns,<a 
name="line.176"></a>
-<span class="sourceLineNo">177</span>      Class&lt;? extends TableMap&gt; 
mapper,<a name="line.177"></a>
-<span class="sourceLineNo">178</span>      Class&lt;?&gt; outputKeyClass,<a 
name="line.178"></a>
-<span class="sourceLineNo">179</span>      Class&lt;?&gt; outputValueClass, 
JobConf job,<a name="line.179"></a>
-<span class="sourceLineNo">180</span>      boolean addDependencyJars, Path 
tmpRestoreDir)<a name="line.180"></a>
-<span class="sourceLineNo">181</span>  throws IOException {<a 
name="line.181"></a>
-<span class="sourceLineNo">182</span>    
TableSnapshotInputFormat.setInput(job, snapshotName, tmpRestoreDir);<a 
name="line.182"></a>
-<span class="sourceLineNo">183</span>    initTableMapJob(snapshotName, 
columns, mapper, outputKeyClass, outputValueClass, job,<a name="line.183"></a>
-<span class="sourceLineNo">184</span>      addDependencyJars, 
TableSnapshotInputFormat.class);<a name="line.184"></a>
-<span class="sourceLineNo">185</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.resetCacheConfig(job);<a 
name="line.185"></a>
-<span class="sourceLineNo">186</span>  }<a name="line.186"></a>
-<span class="sourceLineNo">187</span><a name="line.187"></a>
-<span class="sourceLineNo">188</span>  /**<a name="line.188"></a>
-<span class="sourceLineNo">189</span>   * Use this before submitting a 
TableReduce job. It will<a name="line.189"></a>
-<span class="sourceLineNo">190</span>   * appropriately set up the JobConf.<a 
name="line.190"></a>
-<span class="sourceLineNo">191</span>   *<a name="line.191"></a>
-<span class="sourceLineNo">192</span>   * @param table  The output table.<a 
name="line.192"></a>
-<span class="sourceLineNo">193</span>   * @param reducer  The reducer class to 
use.<a name="line.193"></a>
-<span class="sourceLineNo">194</span>   * @param job  The current job 
configuration to adjust.<a name="line.194"></a>
-<span class="sourceLineNo">195</span>   * @throws IOException When determining 
the region count fails.<a name="line.195"></a>
-<span class="sourceLineNo">196</span>   */<a name="line.196"></a>
-<span class="sourceLineNo">197</span>  public static void 
initTableReduceJob(String table,<a name="line.197"></a>
-<span class="sourceLineNo">198</span>    Class&lt;? extends TableReduce&gt; 
reducer, JobConf job)<a name="line.198"></a>
-<span class="sourceLineNo">199</span>  throws IOException {<a 
name="line.199"></a>
-<span class="sourceLineNo">200</span>    initTableReduceJob(table, reducer, 
job, null);<a name="line.200"></a>
-<span class="sourceLineNo">201</span>  }<a name="line.201"></a>
-<span class="sourceLineNo">202</span><a name="line.202"></a>
-<span class="sourceLineNo">203</span>  /**<a name="line.203"></a>
-<span class="sourceLineNo">204</span>   * Use this before submitting a 
TableReduce job. It will<a name="line.204"></a>
-<span class="sourceLineNo">205</span>   * appropriately set up the JobConf.<a 
name="line.205"></a>
-<span class="sourceLineNo">206</span>   *<a name="line.206"></a>
-<span class="sourceLineNo">207</span>   * @param table  The output table.<a 
name="line.207"></a>
-<span class="sourceLineNo">208</span>   * @param reducer  The reducer class to 
use.<a name="line.208"></a>
-<span class="sourceLineNo">209</span>   * @param job  The current job 
configuration to adjust.<a name="line.209"></a>
-<span class="sourceLineNo">210</span>   * @param partitioner  Partitioner to 
use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.210"></a>
-<span class="sourceLineNo">211</span>   * default partitioner.<a 
name="line.211"></a>
-<span class="sourceLineNo">212</span>   * @throws IOException When determining 
the region count fails.<a name="line.212"></a>
-<span class="sourceLineNo">213</span>   */<a name="line.213"></a>
-<span class="sourceLineNo">214</span>  public static void 
initTableReduceJob(String table,<a name="line.214"></a>
-<span class="sourceLineNo">215</span>    Class&lt;? extends TableReduce&gt; 
reducer, JobConf job, Class partitioner)<a name="line.215"></a>
-<span class="sourceLineNo">216</span>  throws IOException {<a 
name="line.216"></a>
-<span class="sourceLineNo">217</span>    initTableReduceJob(table, reducer, 
job, partitioner, true);<a name="line.217"></a>
-<span class="sourceLineNo">218</span>  }<a name="line.218"></a>
-<span class="sourceLineNo">219</span><a name="line.219"></a>
-<span class="sourceLineNo">220</span>  /**<a name="line.220"></a>
-<span class="sourceLineNo">221</span>   * Use this before submitting a 
TableReduce job. It will<a name="line.221"></a>
-<span class="sourceLineNo">222</span>   * appropriately set up the JobConf.<a 
name="line.222"></a>
-<span class="sourceLineNo">223</span>   *<a name="line.223"></a>
-<span class="sourceLineNo">224</span>   * @param table  The output table.<a 
name="line.224"></a>
-<span class="sourceLineNo">225</span>   * @param reducer  The reducer class to 
use.<a name="line.225"></a>
-<span class="sourceLineNo">226</span>   * @param job  The current job 
configuration to adjust.<a name="line.226"></a>
-<span class="sourceLineNo">227</span>   * @param partitioner  Partitioner to 
use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.227"></a>
-<span class="sourceLineNo">228</span>   * default partitioner.<a 
name="line.228"></a>
-<span class="sourceLineNo">229</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.229"></a>
-<span class="sourceLineNo">230</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.230"></a>
-<span class="sourceLineNo">231</span>   * @throws IOException When determining 
the region count fails.<a name="line.231"></a>
-<span class="sourceLineNo">232</span>   */<a name="line.232"></a>
-<span class="sourceLineNo">233</span>  public static void 
initTableReduceJob(String table,<a name="line.233"></a>
-<span class="sourceLineNo">234</span>    Class&lt;? extends TableReduce&gt; 
reducer, JobConf job, Class partitioner,<a name="line.234"></a>
-<span class="sourceLineNo">235</span>    boolean addDependencyJars) throws 
IOException {<a name="line.235"></a>
-<span class="sourceLineNo">236</span>    
job.setOutputFormat(TableOutputFormat.class);<a name="line.236"></a>
-<span class="sourceLineNo">237</span>    job.setReducerClass(reducer);<a 
name="line.237"></a>
-<span class="sourceLineNo">238</span>    
job.set(TableOutputFormat.OUTPUT_TABLE, table);<a name="line.238"></a>
-<span class="sourceLineNo">239</span>    
job.setOutputKeyClass(ImmutableBytesWritable.class);<a name="line.239"></a>
-<span class="sourceLineNo">240</span>    job.setOutputValueClass(Put.class);<a 
name="line.240"></a>
-<span class="sourceLineNo">241</span>    job.setStrings("io.serializations", 
job.get("io.serializations"),<a name="line.241"></a>
-<span class="sourceLineNo">242</span>        
MutationSerialization.class.getName(), ResultSerialization.class.getName());<a 
name="line.242"></a>
-<span class="sourceLineNo">243</span>    if (partitioner == 
HRegionPartitioner.class) {<a name="line.243"></a>
-<span class="sourceLineNo">244</span>      
job.setPartitionerClass(HRegionPartitioner.class);<a name="line.244"></a>
-<span class="sourceLineNo">245</span>      int regions =<a name="line.245"></a>
-<span class="sourceLineNo">246</span>        
MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), 
TableName.valueOf(table));<a name="line.246"></a>
-<span class="sourceLineNo">247</span>      if (job.getNumReduceTasks() &gt; 
regions) {<a name="line.247"></a>
-<span class="sourceLineNo">248</span>        job.setNumReduceTasks(regions);<a 
name="line.248"></a>
-<span class="sourceLineNo">249</span>      }<a name="line.249"></a>
-<span class="sourceLineNo">250</span>    } else if (partitioner != null) {<a 
name="line.250"></a>
-<span class="sourceLineNo">251</span>      
job.setPartitionerClass(partitioner);<a name="line.251"></a>
-<span class="sourceLineNo">252</span>    }<a name="line.252"></a>
-<span class="sourceLineNo">253</span>    if (addDependencyJars) {<a 
name="line.253"></a>
-<span class="sourceLineNo">254</span>      addDependencyJars(job);<a 
name="line.254"></a>
-<span class="sourceLineNo">255</span>    }<a name="line.255"></a>
-<span class="sourceLineNo">256</span>    initCredentials(job);<a 
name="line.256"></a>
-<span class="sourceLineNo">257</span>  }<a name="line.257"></a>
-<span class="sourceLineNo">258</span><a name="line.258"></a>
-<span class="sourceLineNo">259</span>  public static void 
initCredentials(JobConf job) throws IOException {<a name="line.259"></a>
-<span class="sourceLineNo">260</span>    UserProvider userProvider = 
UserProvider.instantiate(job);<a name="line.260"></a>
-<span class="sourceLineNo">261</span>    if 
(userProvider.isHadoopSecurityEnabled()) {<a name="line.261"></a>
-<span class="sourceLineNo">262</span>      // propagate delegation related 
props from launcher job to MR job<a name="line.262"></a>
-<span class="sourceLineNo">263</span>      if 
(System.getenv("HADOOP_TOKEN_FILE_LOCATION") != null) {<a name="line.263"></a>
-<span class="sourceLineNo">264</span>        
job.set("mapreduce.job.credentials.binary", 
System.getenv("HADOOP_TOKEN_FILE_LOCATION"));<a name="line.264"></a>
-<span class="sourceLineNo">265</span>      }<a name="line.265"></a>
-<span class="sourceLineNo">266</span>    }<a name="line.266"></a>
-<span class="sourceLineNo">267</span><a name="line.267"></a>
-<span class="sourceLineNo">268</span>    if 
(userProvider.isHBaseSecurityEnabled()) {<a name="line.268"></a>
-<span class="sourceLineNo">269</span>      Connection conn = 
ConnectionFactory.createConnection(job);<a name="line.269"></a>
-<span class="sourceLineNo">270</span>      try {<a name="line.270"></a>
-<span class="sourceLineNo">271</span>        // login the server principal (if 
using secure Hadoop)<a name="line.271"></a>
-<span class="sourceLineNo">272</span>        User user = 
userProvider.getCurrent();<a name="line.272"></a>
-<span class="sourceLineNo">273</span>        TokenUtil.addTokenForJob(conn, 
job, user);<a name="line.273"></a>
-<span class="sourceLineNo">274</span>      } catch (InterruptedException ie) 
{<a name="line.274"></a>
-<span class="sourceLineNo">275</span>        ie.printStackTrace();<a 
name="line.275"></a>
-<span class="sourceLineNo">276</span>        
Thread.currentThread().interrupt();<a name="line.276"></a>
-<span class="sourceLineNo">277</span>      } finally {<a name="line.277"></a>
-<span class="sourceLineNo">278</span>        conn.close();<a 
name="line.278"></a>
-<span class="sourceLineNo">279</span>      }<a name="line.279"></a>
-<span class="sourceLineNo">280</span>    }<a name="line.280"></a>
-<span class="sourceLineNo">281</span>  }<a name="line.281"></a>
-<span class="sourceLineNo">282</span><a name="line.282"></a>
-<span class="sourceLineNo">283</span>  /**<a name="line.283"></a>
-<span class="sourceLineNo">284</span>   * Ensures that the given number of 
reduce tasks for the given job<a name="line.284"></a>
-<span class="sourceLineNo">285</span>   * configuration does not exceed the 
number of regions for the given table.<a name="line.285"></a>
-<span class="sourceLineNo">286</span>   *<a name="line.286"></a>
-<span class="sourceLineNo">287</span>   * @param table  The table to get the 
region count for.<a name="line.287"></a>
-<span class="sourceLineNo">288</span>   * @param job  The current job 
configuration to adjust.<a name="line.288"></a>
-<span class="sourceLineNo">289</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.289"></a>
-<span class="sourceLineNo">290</span>   */<a name="line.290"></a>
-<span class="sourceLineNo">291</span>  // Used by tests.<a name="line.291"></a>
-<span class="sourceLineNo">292</span>  public static void 
limitNumReduceTasks(String table, JobConf job)<a name="line.292"></a>
-<span class="sourceLineNo">293</span>  throws IOException {<a 
name="line.293"></a>
-<span class="sourceLineNo">294</span>    int regions =<a name="line.294"></a>
-<span class="sourceLineNo">295</span>      
MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), 
TableName.valueOf(table));<a name="line.295"></a>
-<span class="sourceLineNo">296</span>    if (job.getNumReduceTasks() &gt; 
regions)<a name="line.296"></a>
-<span class="sourceLineNo">297</span>      job.setNumReduceTasks(regions);<a 
name="line.297"></a>
-<span class="sourceLineNo">298</span>  }<a name="line.298"></a>
-<span class="sourceLineNo">299</span><a name="line.299"></a>
-<span class="sourceLineNo">300</span>  /**<a name="line.300"></a>
-<span class="sourceLineNo">301</span>   * Ensures that the given number of map 
tasks for the given job<a name="line.301"></a>
-<span class="sourceLineNo">302</span>   * configuration does not exceed the 
number of regions for the given table.<a name="line.302"></a>
-<span class="sourceLineNo">303</span>   *<a name="line.303"></a>
-<span class="sourceLineNo">304</span>   * @param table  The table to get the 
region count for.<a name="line.304"></a>
-<span class="sourceLineNo">305</span>   * @param job  The current job 
configuration to adjust.<a name="line.305"></a>
-<span class="sourceLineNo">306</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.306"></a>
-<span class="sourceLineNo">307</span>   */<a name="line.307"></a>
-<span class="sourceLineNo">308</span>  // Used by tests.<a name="line.308"></a>
-<span class="sourceLineNo">309</span>  public static void 
limitNumMapTasks(String table, JobConf job)<a name="line.309"></a>
-<span class="sourceLineNo">310</span>  throws IOException {<a 
name="line.310"></a>
-<span class="sourceLineNo">311</span>    int regions =<a name="line.311"></a>
-<span class="sourceLineNo">312</span>      
MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), 
TableName.valueOf(table));<a name="line.312"></a>
-<span class="sourceLineNo">313</span>    if (job.getNumMapTasks() &gt; 
regions)<a name="line.313"></a>
-<span class="sourceLineNo">314</span>      job.setNumMapTasks(regions);<a 
name="line.314"></a>
-<span class="sourceLineNo">315</span>  }<a name="line.315"></a>
-<span class="sourceLineNo">316</span><a name="line.316"></a>
-<span class="sourceLineNo">317</span>  /**<a name="line.317"></a>
-<span class="sourceLineNo">318</span>   * Sets the number of reduce tasks for 
the given job configuration to the<a name="line.318"></a>
-<span class="sourceLineNo">319</span>   * number of regions the given table 
has.<a name="line.319"></a>
-<span class="sourceLineNo">320</span>   *<a name="line.320"></a>
-<span class="sourceLineNo">321</span>   * @param table  The table to get the 
region count for.<a name="line.321"></a>
-<span class="sourceLineNo">322</span>   * @param job  The current job 
configuration to adjust.<a name="line.322"></a>
-<span class="sourceLineNo">323</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.323"></a>
-<span class="sourceLineNo">324</span>   */<a name="line.324"></a>
-<span class="sourceLineNo">325</span>  public static void 
setNumReduceTasks(String table, JobConf job)<a name="line.325"></a>
-<span class="sourceLineNo">326</span>  throws IOException {<a 
name="line.326"></a>
-<span class="sourceLineNo">327</span>    
job.setNumReduceTasks(MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job),<a
 name="line.327"></a>
-<span class="sourceLineNo">328</span>      TableName.valueOf(table)));<a 
name="line.328"></a>
-<span class="sourceLineNo">329</span>  }<a name="line.329"></a>
-<span class="sourceLineNo">330</span><a name="line.330"></a>
-<span class="sourceLineNo">331</span>  /**<a name="line.331"></a>
-<span class="sourceLineNo">332</span>   * Sets the number of map tasks for the 
given job configuration to the<a name="line.332"></a>
-<span class="sourceLineNo">333</span>   * number of regions the given table 
has.<a name="line.333"></a>
-<span class="sourceLineNo">334</span>   *<a name="line.334"></a>
-<span class="sourceLineNo">335</span>   * @param table  The table to get the 
region count for.<a name="line.335"></a>
-<span class="sourceLineNo">336</span>   * @param job  The current job 
configuration to adjust.<a name="line.336"></a>
-<span class="sourceLineNo">337</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.337"></a>
-<span class="sourceLineNo">338</span>   */<a name="line.338"></a>
-<span class="sourceLineNo">339</span>  public static void 
setNumMapTasks(String table, JobConf job)<a name="line.339"></a>
-<span class="sourceLineNo">340</span>  throws IOException {<a 
name="line.340"></a>
-<span class="sourceLineNo">341</span>    
job.setNumMapTasks(MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job),<a
 name="line.341"></a>
-<span class="sourceLineNo">342</span>      TableName.valueOf(table)));<a 
name="line.342"></a>
-<span class="sourceLineNo">343</span>  }<a name="line.343"></a>
-<span class="sourceLineNo">344</span><a name="line.344"></a>
-<span class="sourceLineNo">345</span>  /**<a name="line.345"></a>
-<span class="sourceLineNo">346</span>   * Sets the number of rows to return 
and cache with each scanner iteration.<a name="line.346"></a>
-<span class="sourceLineNo">347</span>   * Higher caching values will enable 
faster mapreduce jobs at the expense of<a name="line.347"></a>
-<span class="sourceLineNo">348</span>   * requiring more heap to contain the 
cached rows.<a name="line.348"></a>
-<span class="sourceLineNo">349</span>   *<a name="line.349"></a>
-<span class="sourceLineNo">350</span>   * @param job The current job 
configuration to adjust.<a name="line.350"></a>
-<span class="sourceLineNo">351</span>   * @param batchSize The number of rows 
to return in batch with each scanner<a name="line.351"></a>
-<span class="sourceLineNo">352</span>   * iteration.<a name="line.352"></a>
-<span class="sourceLineNo">353</span>   */<a name="line.353"></a>
-<span class="sourceLineNo">354</span>  public static void 
setScannerCaching(JobConf job, int batchSize) {<a name="line.354"></a>
-<span class="sourceLineNo">355</span>    
job.setInt("hbase.client.scanner.caching", batchSize);<a name="line.355"></a>
-<span class="sourceLineNo">356</span>  }<a name="line.356"></a>
-<span class="sourceLineNo">357</span><a name="line.357"></a>
-<span class="sourceLineNo">358</span>  /**<a name="line.358"></a>
-<span class="sourceLineNo">359</span>   * @see 
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil#addDependencyJars(org.apache.hadoop.mapreduce.Job)<a
 name="line.359"></a>
-<span class="sourceLineNo">360</span>   */<a name="line.360"></a>
-<span class="sourceLineNo">361</span>  public static void 
addDependencyJars(JobConf job) throws IOException {<a name="line.361"></a>
-<span class="sourceLineNo">362</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.addHBaseDependencyJars(job);<a
 name="line.362"></a>
-<span class="sourceLineNo">363</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.addDependencyJarsForClasses(<a
 name="line.363"></a>
-<span class="sourceLineNo">364</span>      job,<a name="line.364"></a>
-<span class="sourceLineNo">365</span>      job.getMapOutputKeyClass(),<a 
name="line.365"></a>
-<span class="sourceLineNo">366</span>      job.getMapOutputValueClass(),<a 
name="line.366"></a>
-<span class="sourceLineNo">367</span>      job.getOutputKeyClass(),<a 
name="line.367"></a>
-<span class="sourceLineNo">368</span>      job.getOutputValueClass(),<a 
name="line.368"></a>
-<span class="sourceLineNo">369</span>      job.getPartitionerClass(),<a 
name="line.369"></a>
-<span class="sourceLineNo">370</span>      
job.getClass("mapred.input.format.class", TextInputFormat.class, 
InputFormat.class),<a name="line.370"></a>
-<span class="sourceLineNo">371</span>      
job.getClass("mapred.output.format.class", TextOutputFormat.class, 
OutputFormat.class),<a name="line.371"></a>
-<span class="sourceLineNo">372</span>      job.getCombinerClass());<a 
name="line.372"></a>
-<span class="sourceLineNo">373</span>  }<a name="line.373"></a>
-<span class="sourceLineNo">374</span>}<a name="line.374"></a>
+<span class="sourceLineNo">036</span>import 
org.apache.hadoop.hbase.util.RegionSplitter;<a name="line.36"></a>
+<span class="sourceLineNo">037</span>import 
org.apache.hadoop.mapred.FileInputFormat;<a name="line.37"></a>
+<span class="sourceLineNo">038</span>import 
org.apache.hadoop.mapred.InputFormat;<a name="line.38"></a>
+<span class="sourceLineNo">039</span>import 
org.apache.hadoop.mapred.JobConf;<a name="line.39"></a>
+<span class="sourceLineNo">040</span>import 
org.apache.hadoop.mapred.OutputFormat;<a name="line.40"></a>
+<span class="sourceLineNo">041</span>import 
org.apache.hadoop.mapred.TextInputFormat;<a name="line.41"></a>
+<span class="sourceLineNo">042</span>import 
org.apache.hadoop.mapred.TextOutputFormat;<a name="line.42"></a>
+<span class="sourceLineNo">043</span><a name="line.43"></a>
+<span class="sourceLineNo">044</span>import java.io.IOException;<a 
name="line.44"></a>
+<span class="sourceLineNo">045</span>import java.util.Collection;<a 
name="line.45"></a>
+<span class="sourceLineNo">046</span>import java.util.Map;<a 
name="line.46"></a>
+<span class="sourceLineNo">047</span><a name="line.47"></a>
+<span class="sourceLineNo">048</span>/**<a name="line.48"></a>
+<span class="sourceLineNo">049</span> * Utility for {@link TableMap} and 
{@link TableReduce}<a name="line.49"></a>
+<span class="sourceLineNo">050</span> */<a name="line.50"></a>
+<span class="sourceLineNo">051</span>@InterfaceAudience.Public<a 
name="line.51"></a>
+<span class="sourceLineNo">052</span>@SuppressWarnings({ "rawtypes", 
"unchecked" })<a name="line.52"></a>
+<span class="sourceLineNo">053</span>public class TableMapReduceUtil {<a 
name="line.53"></a>
+<span class="sourceLineNo">054</span><a name="line.54"></a>
+<span class="sourceLineNo">055</span>  /**<a name="line.55"></a>
+<span class="sourceLineNo">056</span>   * Use this before submitting a 
TableMap job. It will<a name="line.56"></a>
+<span class="sourceLineNo">057</span>   * appropriately set up the JobConf.<a 
name="line.57"></a>
+<span class="sourceLineNo">058</span>   *<a name="line.58"></a>
+<span class="sourceLineNo">059</span>   * @param table  The table name to read 
from.<a name="line.59"></a>
+<span class="sourceLineNo">060</span>   * @param columns  The columns to 
scan.<a name="line.60"></a>
+<span class="sourceLineNo">061</span>   * @param mapper  The mapper class to 
use.<a name="line.61"></a>
+<span class="sourceLineNo">062</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.62"></a>
+<span class="sourceLineNo">063</span>   * @param outputValueClass  The class 
of the output value.<a name="line.63"></a>
+<span class="sourceLineNo">064</span>   * @param job  The current job 
configuration to adjust.<a name="line.64"></a>
+<span class="sourceLineNo">065</span>   */<a name="line.65"></a>
+<span class="sourceLineNo">066</span>  public static void 
initTableMapJob(String table, String columns,<a name="line.66"></a>
+<span class="sourceLineNo">067</span>    Class&lt;? extends TableMap&gt; 
mapper,<a name="line.67"></a>
+<span class="sourceLineNo">068</span>    Class&lt;?&gt; outputKeyClass,<a 
name="line.68"></a>
+<span class="sourceLineNo">069</span>    Class&lt;?&gt; outputValueClass, 
JobConf job) {<a name="line.69"></a>
+<span class="sourceLineNo">070</span>    initTableMapJob(table, columns, 
mapper, outputKeyClass, outputValueClass, job,<a name="line.70"></a>
+<span class="sourceLineNo">071</span>      true, TableInputFormat.class);<a 
name="line.71"></a>
+<span class="sourceLineNo">072</span>  }<a name="line.72"></a>
+<span class="sourceLineNo">073</span><a name="line.73"></a>
+<span class="sourceLineNo">074</span>  public static void 
initTableMapJob(String table, String columns,<a name="line.74"></a>
+<span class="sourceLineNo">075</span>    Class&lt;? extends TableMap&gt; 
mapper,<a name="line.75"></a>
+<span class="sourceLineNo">076</span>    Class&lt;?&gt; outputKeyClass,<a 
name="line.76"></a>
+<span class="sourceLineNo">077</span>    Class&lt;?&gt; outputValueClass, 
JobConf job, boolean addDependencyJars) {<a name="line.77"></a>
+<span class="sourceLineNo">078</span>    initTableMapJob(table, columns, 
mapper, outputKeyClass, outputValueClass, job,<a name="line.78"></a>
+<span class="sourceLineNo">079</span>      addDependencyJars, 
TableInputFormat.class);<a name="line.79"></a>
+<span class="sourceLineNo">080</span>  }<a name="line.80"></a>
+<span class="sourceLineNo">081</span><a name="line.81"></a>
+<span class="sourceLineNo">082</span>  /**<a name="line.82"></a>
+<span class="sourceLineNo">083</span>   * Use this before submitting a 
TableMap job. It will<a name="line.83"></a>
+<span class="sourceLineNo">084</span>   * appropriately set up the JobConf.<a 
name="line.84"></a>
+<span class="sourceLineNo">085</span>   *<a name="line.85"></a>
+<span class="sourceLineNo">086</span>   * @param table  The table name to read 
from.<a name="line.86"></a>
+<span class="sourceLineNo">087</span>   * @param columns  The columns to 
scan.<a name="line.87"></a>
+<span class="sourceLineNo">088</span>   * @param mapper  The mapper class to 
use.<a name="line.88"></a>
+<span class="sourceLineNo">089</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.89"></a>
+<span class="sourceLineNo">090</span>   * @param outputValueClass  The class 
of the output value.<a name="line.90"></a>
+<span class="sourceLineNo">091</span>   * @param job  The current job 
configuration to adjust.<a name="line.91"></a>
+<span class="sourceLineNo">092</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.92"></a>
+<span class="sourceLineNo">093</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.93"></a>
+<span class="sourceLineNo">094</span>   */<a name="line.94"></a>
+<span class="sourceLineNo">095</span>  public static void 
initTableMapJob(String table, String columns,<a name="line.95"></a>
+<span class="sourceLineNo">096</span>    Class&lt;? extends TableMap&gt; 
mapper,<a name="line.96"></a>
+<span class="sourceLineNo">097</span>    Class&lt;?&gt; outputKeyClass,<a 
name="line.97"></a>
+<span class="sourceLineNo">098</span>    Class&lt;?&gt; outputValueClass, 
JobConf job, boolean addDependencyJars,<a name="line.98"></a>
+<span class="sourceLineNo">099</span>    Class&lt;? extends InputFormat&gt; 
inputFormat) {<a name="line.99"></a>
+<span class="sourceLineNo">100</span><a name="line.100"></a>
+<span class="sourceLineNo">101</span>    job.setInputFormat(inputFormat);<a 
name="line.101"></a>
+<span class="sourceLineNo">102</span>    
job.setMapOutputValueClass(outputValueClass);<a name="line.102"></a>
+<span class="sourceLineNo">103</span>    
job.setMapOutputKeyClass(outputKeyClass);<a name="line.103"></a>
+<span class="sourceLineNo">104</span>    job.setMapperClass(mapper);<a 
name="line.104"></a>
+<span class="sourceLineNo">105</span>    job.setStrings("io.serializations", 
job.get("io.serializations"),<a name="line.105"></a>
+<span class="sourceLineNo">106</span>        
MutationSerialization.class.getName(), ResultSerialization.class.getName());<a 
name="line.106"></a>
+<span class="sourceLineNo">107</span>    FileInputFormat.addInputPaths(job, 
table);<a name="line.107"></a>
+<span class="sourceLineNo">108</span>    job.set(TableInputFormat.COLUMN_LIST, 
columns);<a name="line.108"></a>
+<span class="sourceLineNo">109</span>    if (addDependencyJars) {<a 
name="line.109"></a>
+<span class="sourceLineNo">110</span>      try {<a name="line.110"></a>
+<span class="sourceLineNo">111</span>        addDependencyJars(job);<a 
name="line.111"></a>
+<span class="sourceLineNo">112</span>      } catch (IOException e) {<a 
name="line.112"></a>
+<span class="sourceLineNo">113</span>        e.printStackTrace();<a 
name="line.113"></a>
+<span class="sourceLineNo">114</span>      }<a name="line.114"></a>
+<span class="sourceLineNo">115</span>    }<a name="line.115"></a>
+<span class="sourceLineNo">116</span>    try {<a name="line.116"></a>
+<span class="sourceLineNo">117</span>      initCredentials(job);<a 
name="line.117"></a>
+<span class="sourceLineNo">118</span>    } catch (IOException ioe) {<a 
name="line.118"></a>
+<span class="sourceLineNo">119</span>      // just spit out the stack trace?  
really?<a name="line.119"></a>
+<span class="sourceLineNo">120</span>      ioe.printStackTrace();<a 
name="line.120"></a>
+<span class="sourceLineNo">121</span>    }<a name="line.121"></a>
+<span class="sourceLineNo">122</span>  }<a name="line.122"></a>
+<span class="sourceLineNo">123</span><a name="line.123"></a>
+<span class="sourceLineNo">124</span>  /**<a name="line.124"></a>
+<span class="sourceLineNo">125</span>   * Sets up the job for reading from one 
or more multiple table snapshots, with one or more scans<a name="line.125"></a>
+<span class="sourceLineNo">126</span>   * per snapshot.<a name="line.126"></a>
+<span class="sourceLineNo">127</span>   * It bypasses hbase servers and read 
directly from snapshot files.<a name="line.127"></a>
+<span class="sourceLineNo">128</span>   *<a name="line.128"></a>
+<span class="sourceLineNo">129</span>   * @param snapshotScans     map of 
snapshot name to scans on that snapshot.<a name="line.129"></a>
+<span class="sourceLineNo">130</span>   * @param mapper            The mapper 
class to use.<a name="line.130"></a>
+<span class="sourceLineNo">131</span>   * @param outputKeyClass    The class 
of the output key.<a name="line.131"></a>
+<span class="sourceLineNo">132</span>   * @param outputValueClass  The class 
of the output value.<a name="line.132"></a>
+<span class="sourceLineNo">133</span>   * @param job               The current 
job to adjust.  Make sure the passed job is<a name="line.133"></a>
+<span class="sourceLineNo">134</span>   *                          carrying 
all necessary HBase configuration.<a name="line.134"></a>
+<span class="sourceLineNo">135</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.135"></a>
+<span class="sourceLineNo">136</span>   *                          job classes 
via the distributed cache (tmpjars).<a name="line.136"></a>
+<span class="sourceLineNo">137</span>   */<a name="line.137"></a>
+<span class="sourceLineNo">138</span>  public static void 
initMultiTableSnapshotMapperJob(Map&lt;String, Collection&lt;Scan&gt;&gt; 
snapshotScans,<a name="line.138"></a>
+<span class="sourceLineNo">139</span>      Class&lt;? extends TableMap&gt; 
mapper, Class&lt;?&gt; outputKeyClass, Class&lt;?&gt; outputValueClass,<a 
name="line.139"></a>
+<span class="sourceLineNo">140</span>      JobConf job, boolean 
addDependencyJars, Path tmpRestoreDir) throws IOException {<a 
name="line.140"></a>
+<span class="sourceLineNo">141</span>    
MultiTableSnapshotInputFormat.setInput(job, snapshotScans, tmpRestoreDir);<a 
name="line.141"></a>
+<span class="sourceLineNo">142</span><a name="line.142"></a>
+<span class="sourceLineNo">143</span>    
job.setInputFormat(MultiTableSnapshotInputFormat.class);<a name="line.143"></a>
+<span class="sourceLineNo">144</span>    if (outputValueClass != null) {<a 
name="line.144"></a>
+<span class="sourceLineNo">145</span>      
job.setMapOutputValueClass(outputValueClass);<a name="line.145"></a>
+<span class="sourceLineNo">146</span>    }<a name="line.146"></a>
+<span class="sourceLineNo">147</span>    if (outputKeyClass != null) {<a 
name="line.147"></a>
+<span class="sourceLineNo">148</span>      
job.setMapOutputKeyClass(outputKeyClass);<a name="line.148"></a>
+<span class="sourceLineNo">149</span>    }<a name="line.149"></a>
+<span class="sourceLineNo">150</span>    job.setMapperClass(mapper);<a 
name="line.150"></a>
+<span class="sourceLineNo">151</span>    if (addDependencyJars) {<a 
name="line.151"></a>
+<span class="sourceLineNo">152</span>      addDependencyJars(job);<a 
name="line.152"></a>
+<span class="sourceLineNo">153</span>    }<a name="line.153"></a>
+<span class="sourceLineNo">154</span><a name="line.154"></a>
+<span class="sourceLineNo">155</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.resetCacheConfig(job);<a 
name="line.155"></a>
+<span class="sourceLineNo">156</span>  }<a name="line.156"></a>
+<span class="sourceLineNo">157</span><a name="line.157"></a>
+<span class="sourceLineNo">158</span>  /**<a name="line.158"></a>
+<span class="sourceLineNo">159</span>   * Sets up the job for reading from a 
table snapshot. It bypasses hbase servers<a name="line.159"></a>
+<span class="sourceLineNo">160</span>   * and read directly from snapshot 
files.<a name="line.160"></a>
+<span class="sourceLineNo">161</span>   *<a name="line.161"></a>
+<span class="sourceLineNo">162</span>   * @param snapshotName The name of the 
snapshot (of a table) to read from.<a name="line.162"></a>
+<span class="sourceLineNo">163</span>   * @param columns  The columns to 
scan.<a name="line.163"></a>
+<span class="sourceLineNo">164</span>   * @param mapper  The mapper class to 
use.<a name="line.164"></a>
+<span class="sourceLineNo">165</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.165"></a>
+<span class="sourceLineNo">166</span>   * @param outputValueClass  The class 
of the output value.<a name="line.166"></a>
+<span class="sourceLineNo">167</span>   * @param job  The current job to 
adjust.  Make sure the passed job is<a name="line.167"></a>
+<span class="sourceLineNo">168</span>   * carrying all necessary HBase 
configuration.<a name="line.168"></a>
+<span class="sourceLineNo">169</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.169"></a>
+<span class="sourceLineNo">170</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.170"></a>
+<span class="sourceLineNo">171</span>   * @param tmpRestoreDir a temporary 
directory to copy the snapshot files into. Current user should<a 
name="line.171"></a>
+<span class="sourceLineNo">172</span>   * have write permissions to this 
directory, and this should not be a subdirectory of rootdir.<a 
name="line.172"></a>
+<span class="sourceLineNo">173</span>   * After the job is finished, restore 
directory can be deleted.<a name="line.173"></a>
+<span class="sourceLineNo">174</span>   * @throws IOException When setting up 
the details fails.<a name="line.174"></a>
+<span class="sourceLineNo">175</span>   * @see TableSnapshotInputFormat<a 
name="line.175"></a>
+<span class="sourceLineNo">176</span>   */<a name="line.176"></a>
+<span class="sourceLineNo">177</span>  public static void 
initTableSnapshotMapJob(String snapshotName, String columns,<a 
name="line.177"></a>
+<span class="sourceLineNo">178</span>      Class&lt;? extends TableMap&gt; 
mapper,<a name="line.178"></a>
+<span class="sourceLineNo">179</span>      Class&lt;?&gt; outputKeyClass,<a 
name="line.179"></a>
+<span class="sourceLineNo">180</span>      Class&lt;?&gt; outputValueClass, 
JobConf job,<a name="line.180"></a>
+<span class="sourceLineNo">181</span>      boolean addDependencyJars, Path 
tmpRestoreDir)<a name="line.181"></a>
+<span class="sourceLineNo">182</span>  throws IOException {<a 
name="line.182"></a>
+<span class="sourceLineNo">183</span>    
TableSnapshotInputFormat.setInput(job, snapshotName, tmpRestoreDir);<a 
name="line.183"></a>
+<span class="sourceLineNo">184</span>    initTableMapJob(snapshotName, 
columns, mapper, outputKeyClass, outputValueClass, job,<a name="line.184"></a>
+<span class="sourceLineNo">185</span>      addDependencyJars, 
TableSnapshotInputFormat.class);<a name="line.185"></a>
+<span class="sourceLineNo">186</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.resetCacheConfig(job);<a 
name="line.186"></a>
+<span class="sourceLineNo">187</span>  }<a name="line.187"></a>
+<span class="sourceLineNo">188</span><a name="line.188"></a>
+<span class="sourceLineNo">189</span>  /**<a name="line.189"></a>
+<span class="sourceLineNo">190</span>   * Sets up the job for reading from a 
table snapshot. It bypasses hbase servers<a name="line.190"></a>
+<span class="sourceLineNo">191</span>   * and read directly from snapshot 
files.<a name="line.191"></a>
+<span class="sourceLineNo">192</span>   *<a name="line.192"></a>
+<span class="sourceLineNo">193</span>   * @param snapshotName The name of the 
snapshot (of a table) to read from.<a name="line.193"></a>
+<span class="sourceLineNo">194</span>   * @param columns  The columns to 
scan.<a name="line.194"></a>
+<span class="sourceLineNo">195</span>   * @param mapper  The mapper class to 
use.<a name="line.195"></a>
+<span class="sourceLineNo">196</span>   * @param outputKeyClass  The class of 
the output key.<a name="line.196"></a>
+<span class="sourceLineNo">197</span>   * @param outputValueClass  The class 
of the output value.<a name="line.197"></a>
+<span class="sourceLineNo">198</span>   * @param jobConf  The current job to 
adjust.  Make sure the passed job is<a name="line.198"></a>
+<span class="sourceLineNo">199</span>   * carrying all necessary HBase 
configuration.<a name="line.199"></a>
+<span class="sourceLineNo">200</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.200"></a>
+<span class="sourceLineNo">201</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.201"></a>
+<span class="sourceLineNo">202</span>   * @param tmpRestoreDir a temporary 
directory to copy the snapshot files into. Current user should<a 
name="line.202"></a>
+<span class="sourceLineNo">203</span>   * have write permissions to this 
directory, and this should not be a subdirectory of rootdir.<a 
name="line.203"></a>
+<span class="sourceLineNo">204</span>   * After the job is finished, restore 
directory can be deleted.<a name="line.204"></a>
+<span class="sourceLineNo">205</span>   * @param splitAlgo algorithm to 
split<a name="line.205"></a>
+<span class="sourceLineNo">206</span>   * @param numSplitsPerRegion how many 
input splits to generate per one region<a name="line.206"></a>
+<span class="sourceLineNo">207</span>   * @throws IOException When setting up 
the details fails.<a name="line.207"></a>
+<span class="sourceLineNo">208</span>   * @see TableSnapshotInputFormat<a 
name="line.208"></a>
+<span class="sourceLineNo">209</span>   */<a name="line.209"></a>
+<span class="sourceLineNo">210</span>  public static void 
initTableSnapshotMapJob(String snapshotName, String columns,<a 
name="line.210"></a>
+<span class="sourceLineNo">211</span>                                          
   Class&lt;? extends TableMap&gt; mapper,<a name="line.211"></a>
+<span class="sourceLineNo">212</span>                                          
   Class&lt;?&gt; outputKeyClass,<a name="line.212"></a>
+<span class="sourceLineNo">213</span>                                          
   Class&lt;?&gt; outputValueClass, JobConf jobConf,<a name="line.213"></a>
+<span class="sourceLineNo">214</span>                                          
   boolean addDependencyJars, Path tmpRestoreDir,<a name="line.214"></a>
+<span class="sourceLineNo">215</span>                                          
   RegionSplitter.SplitAlgorithm splitAlgo,<a name="line.215"></a>
+<span class="sourceLineNo">216</span>                                          
   int numSplitsPerRegion)<a name="line.216"></a>
+<span class="sourceLineNo">217</span>          throws IOException {<a 
name="line.217"></a>
+<span class="sourceLineNo">218</span>    
TableSnapshotInputFormat.setInput(jobConf, snapshotName, tmpRestoreDir, 
splitAlgo,<a name="line.218"></a>
+<span class="sourceLineNo">219</span>            numSplitsPerRegion);<a 
name="line.219"></a>
+<span class="sourceLineNo">220</span>    initTableMapJob(snapshotName, 
columns, mapper, outputKeyClass, outputValueClass, jobConf,<a 
name="line.220"></a>
+<span class="sourceLineNo">221</span>            addDependencyJars, 
TableSnapshotInputFormat.class);<a name="line.221"></a>
+<span class="sourceLineNo">222</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.resetCacheConfig(jobConf);<a
 name="line.222"></a>
+<span class="sourceLineNo">223</span>  }<a name="line.223"></a>
+<span class="sourceLineNo">224</span><a name="line.224"></a>
+<span class="sourceLineNo">225</span><a name="line.225"></a>
+<span class="sourceLineNo">226</span>  /**<a name="line.226"></a>
+<span class="sourceLineNo">227</span>   * Use this before submitting a 
TableReduce job. It will<a name="line.227"></a>
+<span class="sourceLineNo">228</span>   * appropriately set up the JobConf.<a 
name="line.228"></a>
+<span class="sourceLineNo">229</span>   *<a name="line.229"></a>
+<span class="sourceLineNo">230</span>   * @param table  The output table.<a 
name="line.230"></a>
+<span class="sourceLineNo">231</span>   * @param reducer  The reducer class to 
use.<a name="line.231"></a>
+<span class="sourceLineNo">232</span>   * @param job  The current job 
configuration to adjust.<a name="line.232"></a>
+<span class="sourceLineNo">233</span>   * @throws IOException When determining 
the region count fails.<a name="line.233"></a>
+<span class="sourceLineNo">234</span>   */<a name="line.234"></a>
+<span class="sourceLineNo">235</span>  public static void 
initTableReduceJob(String table,<a name="line.235"></a>
+<span class="sourceLineNo">236</span>    Class&lt;? extends TableReduce&gt; 
reducer, JobConf job)<a name="line.236"></a>
+<span class="sourceLineNo">237</span>  throws IOException {<a 
name="line.237"></a>
+<span class="sourceLineNo">238</span>    initTableReduceJob(table, reducer, 
job, null);<a name="line.238"></a>
+<span class="sourceLineNo">239</span>  }<a name="line.239"></a>
+<span class="sourceLineNo">240</span><a name="line.240"></a>
+<span class="sourceLineNo">241</span>  /**<a name="line.241"></a>
+<span class="sourceLineNo">242</span>   * Use this before submitting a 
TableReduce job. It will<a name="line.242"></a>
+<span class="sourceLineNo">243</span>   * appropriately set up the JobConf.<a 
name="line.243"></a>
+<span class="sourceLineNo">244</span>   *<a name="line.244"></a>
+<span class="sourceLineNo">245</span>   * @param table  The output table.<a 
name="line.245"></a>
+<span class="sourceLineNo">246</span>   * @param reducer  The reducer class to 
use.<a name="line.246"></a>
+<span class="sourceLineNo">247</span>   * @param job  The current job 
configuration to adjust.<a name="line.247"></a>
+<span class="sourceLineNo">248</span>   * @param partitioner  Partitioner to 
use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.248"></a>
+<span class="sourceLineNo">249</span>   * default partitioner.<a 
name="line.249"></a>
+<span class="sourceLineNo">250</span>   * @throws IOException When determining 
the region count fails.<a name="line.250"></a>
+<span class="sourceLineNo">251</span>   */<a name="line.251"></a>
+<span class="sourceLineNo">252</span>  public static void 
initTableReduceJob(String table,<a name="line.252"></a>
+<span class="sourceLineNo">253</span>    Class&lt;? extends TableReduce&gt; 
reducer, JobConf job, Class partitioner)<a name="line.253"></a>
+<span class="sourceLineNo">254</span>  throws IOException {<a 
name="line.254"></a>
+<span class="sourceLineNo">255</span>    initTableReduceJob(table, reducer, 
job, partitioner, true);<a name="line.255"></a>
+<span class="sourceLineNo">256</span>  }<a name="line.256"></a>
+<span class="sourceLineNo">257</span><a name="line.257"></a>
+<span class="sourceLineNo">258</span>  /**<a name="line.258"></a>
+<span class="sourceLineNo">259</span>   * Use this before submitting a 
TableReduce job. It will<a name="line.259"></a>
+<span class="sourceLineNo">260</span>   * appropriately set up the JobConf.<a 
name="line.260"></a>
+<span class="sourceLineNo">261</span>   *<a name="line.261"></a>
+<span class="sourceLineNo">262</span>   * @param table  The output table.<a 
name="line.262"></a>
+<span class="sourceLineNo">263</span>   * @param reducer  The reducer class to 
use.<a name="line.263"></a>
+<span class="sourceLineNo">264</span>   * @param job  The current job 
configuration to adjust.<a name="line.264"></a>
+<span class="sourceLineNo">265</span>   * @param partitioner  Partitioner to 
use. Pass &lt;code&gt;null&lt;/code&gt; to use<a name="line.265"></a>
+<span class="sourceLineNo">266</span>   * default partitioner.<a 
name="line.266"></a>
+<span class="sourceLineNo">267</span>   * @param addDependencyJars upload 
HBase jars and jars for any of the configured<a name="line.267"></a>
+<span class="sourceLineNo">268</span>   *           job classes via the 
distributed cache (tmpjars).<a name="line.268"></a>
+<span class="sourceLineNo">269</span>   * @throws IOException When determining 
the region count fails.<a name="line.269"></a>
+<span class="sourceLineNo">270</span>   */<a name="line.270"></a>
+<span class="sourceLineNo">271</span>  public static void 
initTableReduceJob(String table,<a name="line.271"></a>
+<span class="sourceLineNo">272</span>    Class&lt;? extends TableReduce&gt; 
reducer, JobConf job, Class partitioner,<a name="line.272"></a>
+<span class="sourceLineNo">273</span>    boolean addDependencyJars) throws 
IOException {<a name="line.273"></a>
+<span class="sourceLineNo">274</span>    
job.setOutputFormat(TableOutputFormat.class);<a name="line.274"></a>
+<span class="sourceLineNo">275</span>    job.setReducerClass(reducer);<a 
name="line.275"></a>
+<span class="sourceLineNo">276</span>    
job.set(TableOutputFormat.OUTPUT_TABLE, table);<a name="line.276"></a>
+<span class="sourceLineNo">277</span>    
job.setOutputKeyClass(ImmutableBytesWritable.class);<a name="line.277"></a>
+<span class="sourceLineNo">278</span>    job.setOutputValueClass(Put.class);<a 
name="line.278"></a>
+<span class="sourceLineNo">279</span>    job.setStrings("io.serializations", 
job.get("io.serializations"),<a name="line.279"></a>
+<span class="sourceLineNo">280</span>        
MutationSerialization.class.getName(), ResultSerialization.class.getName());<a 
name="line.280"></a>
+<span class="sourceLineNo">281</span>    if (partitioner == 
HRegionPartitioner.class) {<a name="line.281"></a>
+<span class="sourceLineNo">282</span>      
job.setPartitionerClass(HRegionPartitioner.class);<a name="line.282"></a>
+<span class="sourceLineNo">283</span>      int regions =<a name="line.283"></a>
+<span class="sourceLineNo">284</span>        
MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), 
TableName.valueOf(table));<a name="line.284"></a>
+<span class="sourceLineNo">285</span>      if (job.getNumReduceTasks() &gt; 
regions) {<a name="line.285"></a>
+<span class="sourceLineNo">286</span>        job.setNumReduceTasks(regions);<a 
name="line.286"></a>
+<span class="sourceLineNo">287</span>      }<a name="line.287"></a>
+<span class="sourceLineNo">288</span>    } else if (partitioner != null) {<a 
name="line.288"></a>
+<span class="sourceLineNo">289</span>      
job.setPartitionerClass(partitioner);<a name="line.289"></a>
+<span class="sourceLineNo">290</span>    }<a name="line.290"></a>
+<span class="sourceLineNo">291</span>    if (addDependencyJars) {<a 
name="line.291"></a>
+<span class="sourceLineNo">292</span>      addDependencyJars(job);<a 
name="line.292"></a>
+<span class="sourceLineNo">293</span>    }<a name="line.293"></a>
+<span class="sourceLineNo">294</span>    initCredentials(job);<a 
name="line.294"></a>
+<span class="sourceLineNo">295</span>  }<a name="line.295"></a>
+<span class="sourceLineNo">296</span><a name="line.296"></a>
+<span class="sourceLineNo">297</span>  public static void 
initCredentials(JobConf job) throws IOException {<a name="line.297"></a>
+<span class="sourceLineNo">298</span>    UserProvider userProvider = 
UserProvider.instantiate(job);<a name="line.298"></a>
+<span class="sourceLineNo">299</span>    if 
(userProvider.isHadoopSecurityEnabled()) {<a name="line.299"></a>
+<span class="sourceLineNo">300</span>      // propagate delegation related 
props from launcher job to MR job<a name="line.300"></a>
+<span class="sourceLineNo">301</span>      if 
(System.getenv("HADOOP_TOKEN_FILE_LOCATION") != null) {<a name="line.301"></a>
+<span class="sourceLineNo">302</span>        
job.set("mapreduce.job.credentials.binary", 
System.getenv("HADOOP_TOKEN_FILE_LOCATION"));<a name="line.302"></a>
+<span class="sourceLineNo">303</span>      }<a name="line.303"></a>
+<span class="sourceLineNo">304</span>    }<a name="line.304"></a>
+<span class="sourceLineNo">305</span><a name="line.305"></a>
+<span class="sourceLineNo">306</span>    if 
(userProvider.isHBaseSecurityEnabled()) {<a name="line.306"></a>
+<span class="sourceLineNo">307</span>      Connection conn = 
ConnectionFactory.createConnection(job);<a name="line.307"></a>
+<span class="sourceLineNo">308</span>      try {<a name="line.308"></a>
+<span class="sourceLineNo">309</span>        // login the server principal (if 
using secure Hadoop)<a name="line.309"></a>
+<span class="sourceLineNo">310</span>        User user = 
userProvider.getCurrent();<a name="line.310"></a>
+<span class="sourceLineNo">311</span>        TokenUtil.addTokenForJob(conn, 
job, user);<a name="line.311"></a>
+<span class="sourceLineNo">312</span>      } catch (InterruptedException ie) 
{<a name="line.312"></a>
+<span class="sourceLineNo">313</span>        ie.printStackTrace();<a 
name="line.313"></a>
+<span class="sourceLineNo">314</span>        
Thread.currentThread().interrupt();<a name="line.314"></a>
+<span class="sourceLineNo">315</span>      } finally {<a name="line.315"></a>
+<span class="sourceLineNo">316</span>        conn.close();<a 
name="line.316"></a>
+<span class="sourceLineNo">317</span>      }<a name="line.317"></a>
+<span class="sourceLineNo">318</span>    }<a name="line.318"></a>
+<span class="sourceLineNo">319</span>  }<a name="line.319"></a>
+<span class="sourceLineNo">320</span><a name="line.320"></a>
+<span class="sourceLineNo">321</span>  /**<a name="line.321"></a>
+<span class="sourceLineNo">322</span>   * Ensures that the given number of 
reduce tasks for the given job<a name="line.322"></a>
+<span class="sourceLineNo">323</span>   * configuration does not exceed the 
number of regions for the given table.<a name="line.323"></a>
+<span class="sourceLineNo">324</span>   *<a name="line.324"></a>
+<span class="sourceLineNo">325</span>   * @param table  The table to get the 
region count for.<a name="line.325"></a>
+<span class="sourceLineNo">326</span>   * @param job  The current job 
configuration to adjust.<a name="line.326"></a>
+<span class="sourceLineNo">327</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.327"></a>
+<span class="sourceLineNo">328</span>   */<a name="line.328"></a>
+<span class="sourceLineNo">329</span>  // Used by tests.<a name="line.329"></a>
+<span class="sourceLineNo">330</span>  public static void 
limitNumReduceTasks(String table, JobConf job)<a name="line.330"></a>
+<span class="sourceLineNo">331</span>  throws IOException {<a 
name="line.331"></a>
+<span class="sourceLineNo">332</span>    int regions =<a name="line.332"></a>
+<span class="sourceLineNo">333</span>      
MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), 
TableName.valueOf(table));<a name="line.333"></a>
+<span class="sourceLineNo">334</span>    if (job.getNumReduceTasks() &gt; 
regions)<a name="line.334"></a>
+<span class="sourceLineNo">335</span>      job.setNumReduceTasks(regions);<a 
name="line.335"></a>
+<span class="sourceLineNo">336</span>  }<a name="line.336"></a>
+<span class="sourceLineNo">337</span><a name="line.337"></a>
+<span class="sourceLineNo">338</span>  /**<a name="line.338"></a>
+<span class="sourceLineNo">339</span>   * Ensures that the given number of map 
tasks for the given job<a name="line.339"></a>
+<span class="sourceLineNo">340</span>   * configuration does not exceed the 
number of regions for the given table.<a name="line.340"></a>
+<span class="sourceLineNo">341</span>   *<a name="line.341"></a>
+<span class="sourceLineNo">342</span>   * @param table  The table to get the 
region count for.<a name="line.342"></a>
+<span class="sourceLineNo">343</span>   * @param job  The current job 
configuration to adjust.<a name="line.343"></a>
+<span class="sourceLineNo">344</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.344"></a>
+<span class="sourceLineNo">345</span>   */<a name="line.345"></a>
+<span class="sourceLineNo">346</span>  // Used by tests.<a name="line.346"></a>
+<span class="sourceLineNo">347</span>  public static void 
limitNumMapTasks(String table, JobConf job)<a name="line.347"></a>
+<span class="sourceLineNo">348</span>  throws IOException {<a 
name="line.348"></a>
+<span class="sourceLineNo">349</span>    int regions =<a name="line.349"></a>
+<span class="sourceLineNo">350</span>      
MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), 
TableName.valueOf(table));<a name="line.350"></a>
+<span class="sourceLineNo">351</span>    if (job.getNumMapTasks() &gt; 
regions)<a name="line.351"></a>
+<span class="sourceLineNo">352</span>      job.setNumMapTasks(regions);<a 
name="line.352"></a>
+<span class="sourceLineNo">353</span>  }<a name="line.353"></a>
+<span class="sourceLineNo">354</span><a name="line.354"></a>
+<span class="sourceLineNo">355</span>  /**<a name="line.355"></a>
+<span class="sourceLineNo">356</span>   * Sets the number of reduce tasks for 
the given job configuration to the<a name="line.356"></a>
+<span class="sourceLineNo">357</span>   * number of regions the given table 
has.<a name="line.357"></a>
+<span class="sourceLineNo">358</span>   *<a name="line.358"></a>
+<span class="sourceLineNo">359</span>   * @param table  The table to get the 
region count for.<a name="line.359"></a>
+<span class="sourceLineNo">360</span>   * @param job  The current job 
configuration to adjust.<a name="line.360"></a>
+<span class="sourceLineNo">361</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.361"></a>
+<span class="sourceLineNo">362</span>   */<a name="line.362"></a>
+<span class="sourceLineNo">363</span>  public static void 
setNumReduceTasks(String table, JobConf job)<a name="line.363"></a>
+<span class="sourceLineNo">364</span>  throws IOException {<a 
name="line.364"></a>
+<span class="sourceLineNo">365</span>    
job.setNumReduceTasks(MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job),<a
 name="line.365"></a>
+<span class="sourceLineNo">366</span>      TableName.valueOf(table)));<a 
name="line.366"></a>
+<span class="sourceLineNo">367</span>  }<a name="line.367"></a>
+<span class="sourceLineNo">368</span><a name="line.368"></a>
+<span class="sourceLineNo">369</span>  /**<a name="line.369"></a>
+<span class="sourceLineNo">370</span>   * Sets the number of map tasks for the 
given job configuration to the<a name="line.370"></a>
+<span class="sourceLineNo">371</span>   * number of regions the given table 
has.<a name="line.371"></a>
+<span class="sourceLineNo">372</span>   *<a name="line.372"></a>
+<span class="sourceLineNo">373</span>   * @param table  The table to get the 
region count for.<a name="line.373"></a>
+<span class="sourceLineNo">374</span>   * @param job  The current job 
configuration to adjust.<a name="line.374"></a>
+<span class="sourceLineNo">375</span>   * @throws IOException When retrieving 
the table details fails.<a name="line.375"></a>
+<span class="sourceLineNo">376</span>   */<a name="line.376"></a>
+<span class="sourceLineNo">377</span>  public static void 
setNumMapTasks(String table, JobConf job)<a name="line.377"></a>
+<span class="sourceLineNo">378</span>  throws IOException {<a 
name="line.378"></a>
+<span class="sourceLineNo">379</span>    
job.setNumMapTasks(MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job),<a
 name="line.379"></a>
+<span class="sourceLineNo">380</span>      TableName.valueOf(table)));<a 
name="line.380"></a>
+<span class="sourceLineNo">381</span>  }<a name="line.381"></a>
+<span class="sourceLineNo">382</span><a name="line.382"></a>
+<span class="sourceLineNo">383</span>  /**<a name="line.383"></a>
+<span class="sourceLineNo">384</span>   * Sets the number of rows to return 
and cache with each scanner iteration.<a name="line.384"></a>
+<span class="sourceLineNo">385</span>   * Higher caching values will enable 
faster mapreduce jobs at the expense of<a name="line.385"></a>
+<span class="sourceLineNo">386</span>   * requiring more heap to contain the 
cached rows.<a name="line.386"></a>
+<span class="sourceLineNo">387</span>   *<a name="line.387"></a>
+<span class="sourceLineNo">388</span>   * @param job The current job 
configuration to adjust.<a name="line.388"></a>
+<span class="sourceLineNo">389</span>   * @param batchSize The number of rows 
to return in batch with each scanner<a name="line.389"></a>
+<span class="sourceLineNo">390</span>   * iteration.<a name="line.390"></a>
+<span class="sourceLineNo">391</span>   */<a name="line.391"></a>
+<span class="sourceLineNo">392</span>  public static void 
setScannerCaching(JobConf job, int batchSize) {<a name="line.392"></a>
+<span class="sourceLineNo">393</span>    
job.setInt("hbase.client.scanner.caching", batchSize);<a name="line.393"></a>
+<span class="sourceLineNo">394</span>  }<a name="line.394"></a>
+<span class="sourceLineNo">395</span><a name="line.395"></a>
+<span class="sourceLineNo">396</span>  /**<a name="line.396"></a>
+<span class="sourceLineNo">397</span>   * @see 
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil#addDependencyJars(org.apache.hadoop.mapreduce.Job)<a
 name="line.397"></a>
+<span class="sourceLineNo">398</span>   */<a name="line.398"></a>
+<span class="sourceLineNo">399</span>  public static void 
addDependencyJars(JobConf job) throws IOException {<a name="line.399"></a>
+<span class="sourceLineNo">400</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.addHBaseDependencyJars(job);<a
 name="line.400"></a>
+<span class="sourceLineNo">401</span>    
org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.addDependencyJarsForClasses(<a
 name="line.401"></a>
+<span class="sourceLineNo">402</span>      job,<a name="line.402"></a>
+<span class="sourceLineNo">403</span>      job.getMapOutputKeyClass(),<a 
name="line.403"></a>
+<span class="sourceLineNo">404</span>      job.getMapOutputValueClass(),<a 
name="line.404"></a>
+<span class="sourceLineNo">405</span>      job.getOutputKeyClass(),<a 
name="line.405"></a>
+<span class="sourceLineNo">406</span>      job.getOutputValueClass(),<a 
name="line.406"></a>
+<span class="sourceLineNo">407</span>      job.getPartitionerClass(),<a 
name="line.407"></a>
+<span class="sourceLineNo">408</span>      
job.getClass("mapred.input.format.class", TextInputFormat.class, 
InputFormat.class),<a name="line.408"></a>
+<span class="sourceLineNo">409</span>      
job.getClass("mapred.output.format.class", TextOutputFormat.class, 
OutputFormat.class),<a name="line.409"></a>
+<span class="sourceLineNo">410</span>      job.getCombinerClass());<a 
name="line.410"></a>
+<span class="sourceLineNo">411</span>  }<a name="line.411"></a>
+<span class="sourceLineNo">412</span>}<a name="line.412"></a>
[47/51] [partial] hbase-site git commit: Published site at .

Reply via email to
