ZOOKEEPER-440. update the performance documentation in forrest

git-svn-id: https://svn.apache.org/repos/asf/hadoop/zookeeper/trunk@789507 13f79535-47bb-0310-9956-ffa450edef68

CHANGES.txt

@@ -112,32 +112,35 @@ BUGFIXES:
   ZOOKEEPER-421. zkpython run_tests.sh is missing #!
   (henry robinson via phunt)
 
-  ZOOKEEPER-406. address all findbugs warnings in persistence classes. (phunt et al via breed)
+  ZOOKEEPER-406. address all findbugs warnings in persistence classes.
+  (phunt et al via breed)
 
-  ZOOKEEPER-435. allow "super" admin digest based auth to be configurable (phunt via breed)
+  ZOOKEEPER-435. allow "super" admin digest based auth to be configurable
+  (phunt via breed)
 
   ZOOKEEPER-375. zoo_add_auth only retains most recent auth on re-sync.
-(mahadev)
+  (mahadev)
 
   ZOOKEEPER-433.  getacl on root znode (/) fails. (phunt via mahadev)
 
   ZOOKEEPER-408. address all findbugs warnings in persistence classes.
-(phunt, mahadev, flavio via mahadev)
+  (phunt, mahadev, flavio via mahadev)
 
-  ZOOKEEPER-427. ZooKeeper server unexpectedly high CPU utilisation (Sergey Zhuravlev via breed)
+  ZOOKEEPER-427. ZooKeeper server unexpectedly high CPU utilisation
+  (Sergey Zhuravlev via breed)
 
   ZOOKEEPER-446. some traces of the host auth scheme left (breed via mahadev)
 
   ZOOKEEPER-438.  addauth fails to register auth on new client that's not yet
-connected (breed via mahadev)
+  connected (breed via mahadev)
 
   ZOOKEEPER-448. png files do nto work with forrest. (mahadev)
 
-  ZOOKEEPER-417. stray message problem when changing servers (breed via
-mahadev)
+  ZOOKEEPER-417. stray message problem when changing servers
+  (breed via mahadev)
 
   ZOOKEEPER-449. sesssionmoved in java code and ZCLOSING in C have the same
-value. (mahadev)
+  value. (mahadev)
 
 IMPROVEMENTS:
   ZOOKEEPER-308. improve the atomic broadcast performance 3x.
@@ -215,41 +218,48 @@ IMPROVEMENTS:
   ZOOKEEPER-431. Expose methods to ease ZK integration. (Jean-Daniel via breed)
 
   ZOOKEEPER-396. race condition in zookeeper client library between
-zookeeper_close and zoo_synchronous api. (mahadev)
+  zookeeper_close and zoo_synchronous api. (mahadev)
 
   ZOOKEEPER-196. doxygen comment for state argument of watcher_fn typedef and
-implementation differ ("...one of the *_STATE constants, otherwise -1") (breed
-via mahadev)
+  implementation differ ("...one of the *_STATE constants, otherwise -1")
+  (breed via mahadev)
 
-  ZOOKEEPER-336. single bad client can cause server to stop accepting connections (henry robinson via breed)
+  ZOOKEEPER-336. single bad client can cause server to stop accepting
+  connections (henry robinson via breed)
  
-  ZOOKEEPER-434. the java shell should indicate connection status on command prompt (henry robinson via breed)
+  ZOOKEEPER-434. the java shell should indicate connection status on command
+  prompt (henry robinson via breed)
 
   ZOOKEEPER-437. Variety of Documentation Updates (grant via mahadev)
 
   ZOOKEEPER-443. trace logging in watch notification not wrapped with
-istraceneabled - inefficient (pat via mahadev)
+  istraceneabled - inefficient (pat via mahadev)
 
-  ZOOKEEPER-432. Various improvements to zkpython bindings. (henry via
-mahadev)
+  ZOOKEEPER-432. Various improvements to zkpython bindings.
+  (henry via mahadev)
 
   ZOOKEEPER-428. logging should be makred as warn rathen than error in
-NIOServerCnxn. (phunt via mahadev)
+  NIOServerCnxn. (phunt via mahadev)
 
-  ZOOKEEPER-422. Java CLI should support ephemeral and sequential node creation (henry via breed)
+  ZOOKEEPER-422. Java CLI should support ephemeral and sequential node creation
+  (henry via breed)
 
   ZOOKEEPER-315. add forrest docs for bookkeeper. (flavio via mahadev)
 
   ZOOKEEPER-329. document how to integrate 3rd party authentication into ZK
-server ACLs. (breed via mahadev)
+  server ACLs. (breed via mahadev)
 
-  ZOOKEEPER-356. Masking bookie failure during writes to a ledger (flavio via breed)
+  ZOOKEEPER-356. Masking bookie failure during writes to a ledger
+  (flavio via breed)
 
   ZOOKEEPER-327. document effects (latency) of storing large amounts of data
-in znodes. (breed via mahadev)
+  in znodes. (breed via mahadev)
 
   ZOOKEEPER-264. docs should include a state transition diagram for client
-state (breed via mahadev)
+  state (breed via mahadev)
+
+  ZOOKEEPER-440. update the performance documentation in forrest
+  (breed via phunt)
 
 NEW FEATURES:
 
@@ -270,6 +280,7 @@ NEW FEATURES:
 
   ZOOKEEPER-237. Add a Chroot request (phunt and mahadev)
 
+
 Release 3.1.0 - 2009-02-06
   
 Non-backward compatible changes:

docs/zookeeperOver.html

@@ -553,11 +553,29 @@ document.write("Last Published: " + document.lastModified);
 <tr>
 <td>
           
-            <img alt="" src="images/zkperfRW.jpg">
+            <img alt="" src="images/zkperfRW-3.2.jpg">
           
         </td>
 </tr>
 </table>
+<p>The figure <a href="#fg_zkPerfRW">ZooKeeper Throughput as the Read-Write Ratio Varies</a> is a throughput
+      graph of ZooKeeper release 3.2 running on servers with dual 2Ghz
+      Xeon and two SATA 15K RPM drives.  One drive was used as a
+      dedicated ZooKeeper log device. The snapshots were written to
+      the OS drive. Write requests were 1K writes and the reads were
+      1K reads.  "Servers" indicate the size of the ZooKeeper
+      ensemble, the number of servers that make up the
+      service. Approximately 30 other servers were used to simulate
+      the clients. The ZooKeeper ensemble was configured such that
+      leaders do not allow connections from clients.</p>
+<div class="note">
+<div class="label">Note</div>
+<div class="content">
+<p>In version 3.2 r/w performance improved by ~2x
+      compared to the <a href="http://hadoop.apache.org/zookeeper/docs/r3.1.1/zookeeperOver.html#Performance">previous
+      3.1 release</a>.</p>
+</div>
+</div>
 <p>Benchmarks also indicate that it is reliable, too. <a href="#fg_zkPerfReliability">Reliability in the Presence of Errors</a> shows how a deployment responds to
       various failures. The events marked in the figure are the
       following:</p>
@@ -598,7 +616,7 @@ document.write("Last Published: " + document.lastModified);
 </li>
       
 </ol>
-<a name="N10179"></a><a name="Reliability"></a>
+<a name="N10188"></a><a name="Reliability"></a>
 <h3 class="h4">Reliability</h3>
 <p>To show the behavior of the system over time as
         failures are injected we ran a ZooKeeper service made up of
@@ -627,7 +645,7 @@ document.write("Last Published: " + document.lastModified);
       ZooKeeper takes less than 200ms to elect a new leader. Third, as
       followers recover, ZooKeeper is able to raise throughput again once they
       start processing requests.</p>
-<a name="N10192"></a><a name="The+ZooKeeper+Project"></a>
+<a name="N101A1"></a><a name="The+ZooKeeper+Project"></a>
 <h3 class="h4">The ZooKeeper Project</h3>
 <p>ZooKeeper has been
         <a href="http://wiki.apache.org/hadoop/ZooKeeper/PoweredBy">

src/docs/src/documentation/content/xdocs/zookeeperOver.xml

@@ -358,10 +358,25 @@
 
         <mediaobject>
           <imageobject>
-            <imagedata fileref="images/zkperfRW.jpg" />
+            <imagedata fileref="images/zkperfRW-3.2.jpg" />
           </imageobject>
         </mediaobject>
       </figure>
+      <para>The figure <xref linkend="fg_zkPerfRW"/> is a throughput
+      graph of ZooKeeper release 3.2 running on servers with dual 2Ghz
+      Xeon and two SATA 15K RPM drives.  One drive was used as a
+      dedicated ZooKeeper log device. The snapshots were written to
+      the OS drive. Write requests were 1K writes and the reads were
+      1K reads.  "Servers" indicate the size of the ZooKeeper
+      ensemble, the number of servers that make up the
+      service. Approximately 30 other servers were used to simulate
+      the clients. The ZooKeeper ensemble was configured such that
+      leaders do not allow connections from clients.</para>
+
+      <note><para>In version 3.2 r/w performance improved by ~2x
+      compared to the <ulink
+      url="http://hadoop.apache.org/zookeeper/docs/r3.1.1/zookeeperOver.html#Performance">previous
+      3.1 release</ulink>.</para></note>
 
       <para>Benchmarks also indicate that it is reliable, too. <xref
       linkend="fg_zkPerfReliability" /> shows how a deployment responds to