diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/pi/package.html b/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/pi/package.html
index 55913a9f62e..ad06b0b0592 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/pi/package.html
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/pi/package.html
@@ -62,11 +62,11 @@ <h3>The Bits of &pi;</h3>
         clusters spread over 20 days.</li>
     <li>The remaining 70% was finished over a weekend on <i>Hammer</i>,
         a 30,000-task-capacity cluster, which was also used for the
-        <a href="http://developer.yahoo.net/blogs/hadoop/2009/05/hadoop_sorts_a_petabyte_in_162.html">petabyte sort benchmark</a>.</li>
+        <a href="http://yahoohadoop.tumblr.com/post/98338791001/hadoop_sorts_a_petabyte_in_162.html">petabyte sort benchmark</a>.</li>
     <li>The log files are available
         <a href="https://issues.apache.org/jira/secure/attachment/12408543/1e15log.zip">here</a>.</li>
     <li>The result was posted in
-        <a href="http://developer.yahoo.net/blogs/hadoop/2009/05/hadoop_computes_the_10151st_bi.html">this YDN blog</a>.</li>
+        <a href="http://yahoohadoop.tumblr.com/post/98338598026/hadoop-computes-the-10-15-1st-bit-of-π">this YDN blog</a>.</li>
 
 </ul></li></p>
 <p><li>The second part of Row 15 (<tt>D3611</tt>)