Differences

This shows you the differences between two versions of the page.

--- courses:mapreduce-tutorial:step-27 [2012/01/28 18:34]
straka
+++ courses:mapreduce-tutorial:step-27 [2012/01/31 14:39] (current)
straka
@@ Line 1: / Line 1: @@
-====== MapReduce Tutorial : Custom data types ======
+====== MapReduce Tutorial : Running multiple Hadoop jobs in one source file ======
-An important feature of the Java API is that custom data and format types can be provided. In this step we implement two custom data types.
+The Java API offers possibility to submit multiple Hadoop job in one source file. A job can be submitted either using
+  * [[http://hadoop.apache.org/common/docs/r1.0.0/api/org/apache/hadoop/mapreduce/Job.html#waitForCompletion(boolean)|job.waitForCompletion]] -- the job is submitted and the method waits for it to finish (successfully or not).
+  * [[http://hadoop.apache.org/common/docs/r1.0.0/api/org/apache/hadoop/mapreduce/Job.html#submit()|job.submit]] -- the job is submitted and the method immediately returns. In this case, the state of the submitted job can be accessed using [[http://hadoop.apache.org/common/docs/r1.0.0/api/org/apache/hadoop/mapreduce/Job.html#isComplete()|job.isComplete]] and [[http://hadoop.apache.org/common/docs/r1.0.0/api/org/apache/hadoop/mapreduce/Job.html#isSuccessful()|job.isSuccessful]]
-===== BERIntWritable =====
+===== Exercise 1 =====
+Improve the [[.:step-25#exercise|sorting exercise]] to handle [[.:step-13#nonuniform-data|nonuniform keys distribution]]. As in the [[.:step-13#nonuniform-data|Perl solution]], run two Hadoop jobs (using one Java source file) -- first samples the input and creates separator, second does the real sorting.
-We want to implement BERIntWritable, which is an ''int'' stored in the format of ''pack "w", $num''. Quoting: //The bytes represent an unsigned integer in base 128, most significant digit first, with as few digits as possible. Bit eight (the high bit) is set on each byte except the last.//
+===== Exercise 2 =====
-The new class must implement the [[http://hadoop.apache.org/common/docs/r1.0.0/api/org/apache/hadoop/io/Writable.html|Writable]] interface, i.e., methods ''readFields'' and ''write'':
+Implement the [[.:step-15|K-means clustering exercise]] in Java. Instead of an controlling script, use the Java class itself to execute the Hadoop job as many times as necessary.
-<code java>
-public class BERIntWritable implements Writable {
-  private int value;
-  public void readFields(DataInput in) throws IOException {
+----
-    value = 0;
-    byte next;
+<html>
-    while (((next = in.readByte()) & 0x80) != 0) {
+<table style="width:100%">
-      value = (value << 7) | (next & 0x7F);
+<tr>
-    }
+<td style="text-align:left; width: 33%; "></html>[[step-26|Step 26]]: Compression and job configuration.<html></td>
-    value = (value << 7) | next;
+<td style="text-align:center; width: 33%; "></html>[[.|Overview]]<html></td>
-  }
+<td style="text-align:right; width: 33%; "></html>[[step-28|Step 28]]: Custom data types.<html></td>
+</tr>
-  public void write(DataOutput out) throws IOException {
+</table>
-    int highest_pos = 28;
+</html>
-    while (highest_pos > 0 && (value & (0x7F << highest_pos)) == 0) highest_pos -= 7;
-    while (highest_pos > 0) {
-      out.writeByte(0x80 | ((value >> highest_pos) & 0x7F));
-      highest_pos -= 7;
-    }
-    out.writeByte(value & 0x7F);
-  }
-</code>
-Some accessory methods are also needed in order to work with the value:
-<code java>
-  public int get() { return value; }
-  public void set(int value) { this.value = value; }
-</code>
-===== PairWritable<A, B> =====

[ Back to the navigation ] [ Back to the content ]

Institute of Formal and Applied Linguistics Wiki

Differences