accumulo-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From els...@apache.org
Subject [2/2] accumulo-website git commit: Jekyll build from master:5679ae1
Date Mon, 06 Mar 2017 23:00:51 GMT
Jekyll build from master:5679ae1

Wrote blog post about security and performance.

Closes apache/accumulo-website#8

Signed-off-by: Josh Elser <elserj@apache.org>


Project: http://git-wip-us.apache.org/repos/asf/accumulo-website/repo
Commit: http://git-wip-us.apache.org/repos/asf/accumulo-website/commit/803c95d0
Tree: http://git-wip-us.apache.org/repos/asf/accumulo-website/tree/803c95d0
Diff: http://git-wip-us.apache.org/repos/asf/accumulo-website/diff/803c95d0

Branch: refs/heads/asf-site
Commit: 803c95d0cd3065eead750ad2027bc49f57c11c87
Parents: f183655
Author: Josh Elser <elserj@apache.org>
Authored: Mon Mar 6 17:58:45 2017 -0500
Committer: Josh Elser <elserj@apache.org>
Committed: Mon Mar 6 17:58:45 2017 -0500

----------------------------------------------------------------------
 README.md                                       |   1 -
 .../23/security-performance-implications.html   | 449 +++++++++++++++++++
 feed.xml                                        | 385 ++++++++++++----
 images/blog/201702_security/accumuloFiles.png   | Bin 0 -> 313628 bytes
 images/blog/201702_security/avgFilesTab.png     | Bin 0 -> 288346 bytes
 images/blog/201702_security/figure1.png         | Bin 0 -> 78725 bytes
 images/blog/201702_security/tableRecs.png       | Bin 0 -> 159556 bytes
 images/blog/201702_security/tableRecsInMem.png  | Bin 0 -> 223097 bytes
 images/blog/201702_security/totalIngest.png     | Bin 0 -> 245159 bytes
 index.html                                      |  14 +-
 news/index.html                                 |   7 +
 11 files changed, 754 insertions(+), 102 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/README.md
----------------------------------------------------------------------
diff --git a/README.md b/README.md
index 9d4b0ef..13e7cc0 100644
--- a/README.md
+++ b/README.md
@@ -28,4 +28,3 @@ This can be done easily by invoking the post-commit hook (either by hand, or aut
 Git to invoke the post-commit hook).
 
 `./_devtools/git-hooks/post-commit`
-

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/blog/2017/02/23/security-performance-implications.html
----------------------------------------------------------------------
diff --git a/blog/2017/02/23/security-performance-implications.html b/blog/2017/02/23/security-performance-implications.html
new file mode 100644
index 0000000..fd23824
--- /dev/null
+++ b/blog/2017/02/23/security-performance-implications.html
@@ -0,0 +1,449 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+<!--
+    Licensed to the Apache Software Foundation (ASF) under one or more
+    contributor license agreements.  See the NOTICE file distributed with
+    this work for additional information regarding copyright ownership.
+    The ASF licenses this file to You under the Apache License, Version 2.0
+    (the "License"); you may not use this file except in compliance with
+    the License.  You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.
+-->
+<meta charset="utf-8">
+<meta http-equiv="X-UA-Compatible" content="IE=edge">
+<meta name="viewport" content="width=device-width, initial-scale=1">
+<link href="https://maxcdn.bootstrapcdn.com/bootswatch/3.3.7/paper/bootstrap.min.css" rel="stylesheet" integrity="sha384-awusxf8AUojygHf2+joICySzB780jVvQaVCAt1clU3QsyAitLGul28Qxb2r1e5g+" crossorigin="anonymous">
+<link href="//netdna.bootstrapcdn.com/font-awesome/4.0.3/css/font-awesome.css" rel="stylesheet">
+<link rel="stylesheet" type="text/css" href="https://cdn.datatables.net/v/bs/jq-2.2.3/dt-1.10.12/datatables.min.css">
+<link href="/css/accumulo.css" rel="stylesheet" type="text/css">
+
+<title>Security Performance Implications</title>
+
+<script src="https://ajax.googleapis.com/ajax/libs/jquery/2.2.4/jquery.min.js"></script>
+<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
+<script type="text/javascript" src="https://cdn.datatables.net/v/bs/jq-2.2.3/dt-1.10.12/datatables.min.js"></script>
+<script>
+  // show location of canonical site if not currently on the canonical site
+  $(function() {
+    var host = window.location.host;
+    if (typeof host !== 'undefined' && host !== 'accumulo.apache.org') {
+      $('#non-canonical').show();
+    }
+  });
+
+  $(function() {
+    // decorate section headers with anchors
+    return $("h2, h3, h4, h5, h6").each(function(i, el) {
+      var $el, icon, id;
+      $el = $(el);
+      id = $el.attr('id');
+      icon = '<i class="fa fa-link"></i>';
+      if (id) {
+        return $el.append($("<a />").addClass("header-link").attr("href", "#" + id).html(icon));
+      }
+    });
+  });
+  
+  // configure Google Analytics
+  (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
+  (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
+  m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
+  })(window,document,'script','//www.google-analytics.com/analytics.js','ga');
+
+  if (ga.hasOwnProperty('loaded') && ga.loaded === true) {
+    ga('create', 'UA-50934829-1', 'apache.org');
+    ga('send', 'pageview');
+  }
+</script>
+
+</head>
+<body style="padding-top: 100px">
+
+  <nav class="navbar navbar-default navbar-fixed-top">
+  <div class="container">
+    <div class="navbar-header">
+      <button type="button" class="navbar-toggle" data-toggle="collapse" data-target="#navbar-items">
+        <span class="sr-only">Toggle navigation</span>
+        <span class="icon-bar"></span>
+        <span class="icon-bar"></span>
+        <span class="icon-bar"></span>
+      </button>
+      <a href="/"><img id="nav-logo" alt="Apache Accumulo" class="img-responsive" src="/images/accumulo-logo.png" width="200"/></a>
+    </div>
+    <div class="collapse navbar-collapse" id="navbar-items">
+      <ul class="nav navbar-nav">
+        <li class="nav-link"><a href="/downloads">Download</a></li>
+        <li class="dropdown">
+        <a class="dropdown-toggle" data-toggle="dropdown" href="#">Releases<span class="caret"></span></a>
+        <ul class="dropdown-menu">
+          <li><a href="/release/accumulo-1.8.1/">1.8.1 (Latest)</a></li>
+          <li><a href="/release/accumulo-1.7.2/">1.7.2</a></li>
+          <li><a href="/release/accumulo-1.6.6/">1.6.6</a></li>
+          <li><a href="/release/">Archive</a></li>
+        </ul>
+        </li>
+        <li class="dropdown">
+        <a class="dropdown-toggle" data-toggle="dropdown" href="#">Documentation<span class="caret"></span></a>
+        <ul class="dropdown-menu">
+          <li><a href="/1.8/accumulo_user_manual.html">User Manual (1.8)</a></li>
+          <li><a href="/1.8/apidocs">Javadocs (1.8)</a></li>
+          <li><a href="/1.8/examples">Examples (1.8)</a></li>
+          <li><a href="/features">Features</a></li>
+          <li><a href="/glossary">Glossary</a></li>
+          <li><a href="/external-docs">External Docs</a></li>
+          <li><a href="/docs-archive/">Archive</a></li>
+        </ul>
+        </li>
+        <li class="dropdown">
+        <a class="dropdown-toggle" data-toggle="dropdown" href="#">Community<span class="caret"></span></a>
+        <ul class="dropdown-menu">
+          <li><a href="/get_involved">Get Involved</a></li>
+          <li><a href="/mailing_list">Mailing Lists</a></li>
+          <li><a href="/people">People</a></li>
+          <li><a href="/related-projects">Related Projects</a></li>
+          <li><a href="/contributor/">Contributor Guide</a></li>
+        </ul>
+        </li>
+      </ul>
+      <ul class="nav navbar-nav navbar-right">
+        <li class="dropdown">
+        <a class="dropdown-toggle" data-toggle="dropdown" href="#">Apache Software Foundation<span class="caret"></span></a>
+        <ul class="dropdown-menu">
+          <li><a href="https://www.apache.org">Apache Homepage <i class="fa fa-external-link"></i></a></li>
+          <li><a href="https://www.apache.org/licenses/LICENSE-2.0">License <i class="fa fa-external-link"></i></a></li>
+          <li><a href="https://www.apache.org/foundation/sponsorship">Sponsorship <i class="fa fa-external-link"></i></a></li>
+          <li><a href="https://www.apache.org/security">Security <i class="fa fa-external-link"></i></a></li>
+          <li><a href="https://www.apache.org/foundation/thanks">Thanks <i class="fa fa-external-link"></i></a></li>
+          <li><a href="https://www.apache.org/foundation/policies/conduct">Code of Conduct <i class="fa fa-external-link"></i></a></li>
+        </ul>
+        </li>
+      </ul>
+    </div>
+  </div>
+</nav>
+
+
+  <div class="container">
+    <div class="row">
+      <div class="col-md-12">
+
+        <div id="non-canonical" style="display: none; background-color: #F0E68C; padding-left: 1em;">
+          Visit the official site at: <a href="https://accumulo.apache.org">https://accumulo.apache.org</a>
+        </div>
+        <div id="content">
+          
+          <h1 class="title">Security Performance Implications</h1>
+          
+          <table>
+
+<tr><td><b>Date </b></td><td> &nbsp;&nbsp;23 Feb 2017 </td></tr>
+
+</table>
+<p>
+
+
+<p>The purpose of this two part series was to measure the performance impact of
+various security configurations on a cluster running Apache Accumulo’s
+continuous ingest suite. The tests were performed using Amazon Web
+Services (AWS), Hortonworks Data Platform 2.4 and Accumulo 1.7. Each of
+the five different security settings in Accumulo 1.7 was tested including 
+no security, SSL, and SASL with Kerberos authentication for the three quality 
+of protection levels (auth, auth-int, auth-conf).  KDC was MIT.  HDFS was 
+configured to use Kerberos for authentication and had service level 
+authorization on. Other than that, no other security settings (HTTPS, RPC 
+protection, data transfer encryption, etc) were enabled.  Timely was a 
+separate, single node HDFS/Zookeeper/Accumulo instance.</p>
+
+<h2 id="intro">Intro</h2>
+
+<p>All runs utilized the continuous ingest suite that ships with Accumulo (a
+standard method to measure performance in Accumulo).  It generates random 
+graph data and inserts it into Accumulo, creating
+a long linked list of entries.  Part 1 was run with just continuous ingest.<br />
+Based on the test results, there was a measurable performance impact as each additional security configuration was put in place.</p>
+
+<h2 id="methodology">Methodology</h2>
+
+<p>We ran 5 tests, one for each security configuration.  Each iteration of each test inserted 2 billion entries.  Batch writers were configured with 500K max mem 
+to artificially inflate the overall write overhead. This was performed on a
+small cluster on AWS.</p>
+
+<p>Each test used one of the following security configurations:</p>
+
+<ul>
+  <li>No security - Default</li>
+  <li>Two way SSL</li>
+  <li>Kerberos/SASL with auth
+    <ul>
+      <li>auth is just Kerberos authentication between client and server.  Each end of the RPC definitively knows who the other is.</li>
+    </ul>
+  </li>
+  <li>Kerberos/SASL with auth-int
+    <ul>
+      <li>Builds on auth, also providing message integrity checks of the data going across the wire. You also know that the message you received was not altered.</li>
+    </ul>
+  </li>
+  <li>Kerberos/SASL with auth-conf
+    <ul>
+      <li>Builds on auth-int, also providing confidentiality of the message that was sent to prevent others from reading it (aka wire-encryption).</li>
+    </ul>
+  </li>
+</ul>
+
+<p>For each test, five iterations were run to obtain a min, max, and median
+time elapsed at each security configuration. After each iteration,
+Hadoop, and Zookeeper processes were restarted, Accumulo tables are
+wiped clean and tables are recreated. In addition, pagecache, dentries
+and inodes are dropped by issuing a ‘3’ command on
+/proc/sys/vm/drop_caches to ensure that the OS is not caching things to disk
+that might affect the benchmark. The following sequence was performed 
+between iterations:</p>
+
+<ol>
+  <li>Bring down Accumulo</li>
+  <li>Bring down Zookeeper</li>
+  <li>Bring down Hadoop</li>
+  <li>Run sync command</li>
+  <li>Drop OS cache</li>
+  <li>Bring up Hadoop</li>
+  <li>Bring up Zookeeper</li>
+  <li>Bring up Accumulo</li>
+  <li>Drop tables</li>
+  <li>Create tables</li>
+</ol>
+
+<p>For each iteration, the results were stored, fed into <a href="https://nationalsecurityagency.github.io/timely/">Timely</a>, and viewed with Grafana.
+Since the runs were executed sequentially, the start epochs for each run did not align.
+To mitigate, the entries for each run were inserted 
+with the same relative epoch for convenient comparison in Grafana.</p>
+
+<p>The table configurations for Accumulo remain the same throughout the
+different iterations and security levels. The Accumulo site
+configurations differ only due to the different settings for the
+security level configurations.</p>
+
+<h2 id="environment">Environment</h2>
+
+<p>In order to perform the testing, a small AWS cluster was setup using 14
+hosts on EC2. Two i2.xlarge instances were used as master nodes and eight
+d2.xlarge instances were used for workers. In addition, two c4.4xlarge
+instances were used for ingesters, one m4.2xlarge instance was used for
+Timely, and one m4.xlarge instance was used for Apache Ambari. A logical
+diagram of the setup is depicted below:</p>
+
+<p><img src="/images/blog/201702_security/figure1.png" alt="" width="400px" /></p>
+
+<p>Figure 1 - Cluster Layout, Roles, and Instance Types on AWS.</p>
+
+<p>The types of nodes and their function are given below:</p>
+
+<table id="instance_types" class="table">
+  <thead>
+    <tr>
+      <th style="text-align: left">Node Type</th>
+      <th style="text-align: left">AWS EC2 Type</th>
+      <th style="text-align: left">EC2 Type Details</th>
+      <th style="text-align: left">Quantity</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td style="text-align: left">Ingest Nodes</td>
+      <td style="text-align: left">c4.4xlarge</td>
+      <td style="text-align: left">16 core, 30 GB RAM</td>
+      <td style="text-align: left">2</td>
+    </tr>
+    <tr>
+      <td style="text-align: left">Worker Node</td>
+      <td style="text-align: left">d2.xlarge</td>
+      <td style="text-align: left">4 cores, 30.5 GB RAM, 3x2T GB HD</td>
+      <td style="text-align: left">8</td>
+    </tr>
+    <tr>
+      <td style="text-align: left">Master Node</td>
+      <td style="text-align: left">i2.xlarge</td>
+      <td style="text-align: left">4 cores, 30.5 GB RAM, 1x800GB SSD</td>
+      <td style="text-align: left">2</td>
+    </tr>
+    <tr>
+      <td style="text-align: left">Admin Node</td>
+      <td style="text-align: left">m4.xlarge</td>
+      <td style="text-align: left">4 cores, 16 GB RAM</td>
+      <td style="text-align: left">1</td>
+    </tr>
+    <tr>
+      <td style="text-align: left">Timely Node</td>
+      <td style="text-align: left">m4.2xlarge</td>
+      <td style="text-align: left">8 cores, 32 GB RAM</td>
+      <td style="text-align: left">1</td>
+    </tr>
+  </tbody>
+</table>
+
+<p>Table 1 – AWS Instance Types, Role, Details, and Quantities</p>
+
+<h2 id="results">Results</h2>
+
+<p>The median, max, and min of the milliseconds elapsed
+time of all iterations for each test is displayed below. The percentage change
+columns compare the Median, Max, and Min respectively from the no
+security level to each security configuration (e.g. no security Median
+vs. auth-int Median, no security Max vs. auth-int Max).</p>
+
+<table id="results" class="table">
+  <thead>
+    <tr>
+      <th>Security Level</th>
+      <th style="text-align: right">Median</th>
+      <th style="text-align: right">Standard Deviation</th>
+      <th style="text-align: right">Max</th>
+      <th style="text-align: right">Min</th>
+      <th style="text-align: right">% Change (nosec Median vs. Median)</th>
+      <th style="text-align: right">% Change (nosec Max vs. Max)</th>
+      <th style="text-align: right">% Change (nosec Min vs. Min)</th>
+      <th style="text-align: right">Delta from Previous Level (Median)</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td>no security</td>
+      <td style="text-align: right">7829394</td>
+      <td style="text-align: right">139340</td>
+      <td style="text-align: right">8143035</td>
+      <td style="text-align: right">7764309</td>
+      <td style="text-align: right">0.00%</td>
+      <td style="text-align: right">0.00%</td>
+      <td style="text-align: right">0.00%</td>
+      <td style="text-align: right">0.00%</td>
+    </tr>
+    <tr>
+      <td>ssl</td>
+      <td style="text-align: right">8292760</td>
+      <td style="text-align: right">87012</td>
+      <td style="text-align: right">8464060</td>
+      <td style="text-align: right">8204955</td>
+      <td style="text-align: right">5.92%</td>
+      <td style="text-align: right">3.94%</td>
+      <td style="text-align: right">5.68%</td>
+      <td style="text-align: right">5.92%</td>
+    </tr>
+    <tr>
+      <td>auth</td>
+      <td style="text-align: right">8859552</td>
+      <td style="text-align: right">134109</td>
+      <td style="text-align: right">9047971</td>
+      <td style="text-align: right">8657618</td>
+      <td style="text-align: right">13.16%</td>
+      <td style="text-align: right">11.11%</td>
+      <td style="text-align: right">11.51%</td>
+      <td style="text-align: right">6.83%</td>
+    </tr>
+    <tr>
+      <td>auth-int</td>
+      <td style="text-align: right">9500737</td>
+      <td style="text-align: right">155968</td>
+      <td style="text-align: right">9753424</td>
+      <td style="text-align: right">9282371</td>
+      <td style="text-align: right">21.34%</td>
+      <td style="text-align: right">19.78%</td>
+      <td style="text-align: right">19.55%</td>
+      <td style="text-align: right">7.24%</td>
+    </tr>
+    <tr>
+      <td>auth-conf</td>
+      <td style="text-align: right">9479635</td>
+      <td style="text-align: right">170823</td>
+      <td style="text-align: right">9776580</td>
+      <td style="text-align: right">9282189</td>
+      <td style="text-align: right">21.08%</td>
+      <td style="text-align: right">20.06%</td>
+      <td style="text-align: right">19.55%</td>
+      <td style="text-align: right">-0.22%</td>
+    </tr>
+  </tbody>
+</table>
+
+<p>Table 2 – Summarized Time Elapsed for Each Security Level</p>
+
+<h2 id="plots">Plots</h2>
+
+<p>Below are some snapshots of *stats.out elements via Grafana that were inserted
+into Timely with the same relative start time.  Each graph represents a field 
+in the output generated by <a href="https://github.com/apache/accumulo/blob/1.7/test/src/main/java/org/apache/accumulo/test/continuous/ContinuousStatsCollector.java">ContinuousStatsCollector</a></p>
+
+<h3 id="tablerecshttpsgithubcomapacheaccumuloblob17coresrcmainjavaorgapacheaccumulocoremasterthrifttableinfojaval73"><a href="https://github.com/apache/accumulo/blob/1.7/core/src/main/java/org/apache/accumulo/core/master/thrift/TableInfo.java#L73">TABLE_RECS</a></h3>
+<p>(Number of records in the continuous ingest table.  Down sample=1m, aggregate=avg)</p>
+
+<p><a href="/images/blog/201702_security/tableRecs.png"><img src="/images/blog/201702_security/tableRecs.png" alt="" width="800px" /></a></p>
+
+<h3 id="totalingesthttpsgithubcomapacheaccumuloblob17coresrcmainjavaorgapacheaccumulocoremasterthrifttableinfojaval77"><a href="https://github.com/apache/accumulo/blob/1.7/core/src/main/java/org/apache/accumulo/core/master/thrift/TableInfo.java#L77">TOTAL_INGEST</a></h3>
+<p>(Ingest rate for Accumulo instance.  Down sample=5m, aggregate=avg)</p>
+
+<p><a href="/images/blog/201702_security/totalIngest.png"><img src="/images/blog/201702_security/totalIngest.png" alt="" width="800px" /></a></p>
+
+<h3 id="avgfilestablethttpsgithubcomapacheaccumuloblob17coresrcmainjavaorgapacheaccumulocoreutilstatjaval63"><a href="https://github.com/apache/accumulo/blob/1.7/core/src/main/java/org/apache/accumulo/core/util/Stat.java#L63">AVG_FILES/TABLET</a></h3>
+<p>(Average number of files per Accumulo tablet.  Down sample=1m, aggregate=avg)</p>
+
+<p><a href="/images/blog/201702_security/avgFilesTab.png"><img src="/images/blog/201702_security/avgFilesTab.png" alt="" width="800px" /></a></p>
+
+<h3 id="accumulofileshttpsgithubcomapacheaccumuloblob17testsrcmainjavaorgapacheaccumulotestcontinuouscontinuousstatscollectorjaval127"><a href="https://github.com/apache/accumulo/blob/1.7/test/src/main/java/org/apache/accumulo/test/continuous/ContinuousStatsCollector.java#L127">ACCUMULO_FILES</a></h3>
+<p>(Total number of files for Accumulo.  Down sample=1m, aggregate=avg)</p>
+
+<p><a href="/images/blog/201702_security/accumuloFiles.png"><img src="/images/blog/201702_security/accumuloFiles.png" alt="" width="800px" /></a></p>
+
+<p>As can be seen in the plots above, the different security settings have 
+relatively consistent, discernable median run characteristics.  The big
+dip in each TOTAL_INGEST coincides with a large number of major
+compactions, a rate decrease for TABLE_RECS, and a decrease in 
+AVG_FILES/TABLET.</p>
+
+<h2 id="final-thoughts">Final Thoughts</h2>
+
+<p>The biggest performance 
+hits to run duration median (compared to default security) were ~21% for 
+auth-int and auth-conf.  Interesting to note that SSL’s median run duration was 
+lower than all SASL configs and that auth-conf’s was lower than auth-int. 
+Initial  speculation for these oddities revolved around the 
+<a href="https://github.com/m1ch1/mapkeeper/wiki/Thrift-Java-Servers-Compared">Thrift server</a> 
+implementations, but the Thrift differences will not explain the auth-conf/int 
+disparity since both utilize TThreadPoolServer.  It was certainly unexpected that the 
+addition of wire encryption would yield a faster median run duration.  This result 
+prompted, as a sanity check, sniffing the net traffic (in a contrived example 
+not during a timed run) in both auth-conf and auth-int to ensure that the message 
+contents were actually obfuscated in auth-conf (they were) and not obfuscated in 
+auth-int (they weren’t).</p>
+
+<h2 id="future-work">Future Work</h2>
+
+<p>Part 2 of this series will consist of the same continuous ingest loads and 
+configurations with the addition of a query load on the system.</p>
+
+
+
+<p><strong>View all posts in the <a href="/news">news archive</a></strong></p>
+
+        </div>
+
+        
+<footer>
+
+  <p><a href="https://www.apache.org"><img src="/images/feather-small.gif" alt="Apache Software Foundation" id="asf-logo" height="100" /></a></p>
+
+  <p>Copyright © 2011-2017 The Apache Software Foundation. Licensed under the <a href="https://www.apache.org/licenses/LICENSE-2.0">Apache License, Version 2.0</a>.</p>
+
+</footer>
+
+
+      </div>
+    </div>
+  </div>
+</body>
+</html>

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/feed.xml
----------------------------------------------------------------------
diff --git a/feed.xml b/feed.xml
index 6b10896..db56da6 100644
--- a/feed.xml
+++ b/feed.xml
@@ -6,8 +6,8 @@
 </description>
     <link>https://accumulo.apache.org/</link>
     <atom:link href="https://accumulo.apache.org/feed.xml" rel="self" type="application/rss+xml"/>
-    <pubDate>Mon, 27 Feb 2017 11:22:59 -0500</pubDate>
-    <lastBuildDate>Mon, 27 Feb 2017 11:22:59 -0500</lastBuildDate>
+    <pubDate>Mon, 06 Mar 2017 17:58:36 -0500</pubDate>
+    <lastBuildDate>Mon, 06 Mar 2017 17:58:36 -0500</lastBuildDate>
     <generator>Jekyll v3.3.1</generator>
     
       <item>
@@ -182,6 +182,295 @@ HDFS High-Availability instances, forcing NameNode failover.&lt;/p&gt;
       </item>
     
       <item>
+        <title>Security Performance Implications</title>
+        <description>
+&lt;p&gt;The purpose of this two part series was to measure the performance impact of
+various security configurations on a cluster running Apache Accumulo’s
+continuous ingest suite. The tests were performed using Amazon Web
+Services (AWS), Hortonworks Data Platform 2.4 and Accumulo 1.7. Each of
+the five different security settings in Accumulo 1.7 was tested including 
+no security, SSL, and SASL with Kerberos authentication for the three quality 
+of protection levels (auth, auth-int, auth-conf).  KDC was MIT.  HDFS was 
+configured to use Kerberos for authentication and had service level 
+authorization on. Other than that, no other security settings (HTTPS, RPC 
+protection, data transfer encryption, etc) were enabled.  Timely was a 
+separate, single node HDFS/Zookeeper/Accumulo instance.&lt;/p&gt;
+
+&lt;h2 id=&quot;intro&quot;&gt;Intro&lt;/h2&gt;
+
+&lt;p&gt;All runs utilized the continuous ingest suite that ships with Accumulo (a
+standard method to measure performance in Accumulo).  It generates random 
+graph data and inserts it into Accumulo, creating
+a long linked list of entries.  Part 1 was run with just continuous ingest.&lt;br /&gt;
+Based on the test results, there was a measurable performance impact as each additional security configuration was put in place.&lt;/p&gt;
+
+&lt;h2 id=&quot;methodology&quot;&gt;Methodology&lt;/h2&gt;
+
+&lt;p&gt;We ran 5 tests, one for each security configuration.  Each iteration of each test inserted 2 billion entries.  Batch writers were configured with 500K max mem 
+to artificially inflate the overall write overhead. This was performed on a
+small cluster on AWS.&lt;/p&gt;
+
+&lt;p&gt;Each test used one of the following security configurations:&lt;/p&gt;
+
+&lt;ul&gt;
+  &lt;li&gt;No security - Default&lt;/li&gt;
+  &lt;li&gt;Two way SSL&lt;/li&gt;
+  &lt;li&gt;Kerberos/SASL with auth
+    &lt;ul&gt;
+      &lt;li&gt;auth is just Kerberos authentication between client and server.  Each end of the RPC definitively knows who the other is.&lt;/li&gt;
+    &lt;/ul&gt;
+  &lt;/li&gt;
+  &lt;li&gt;Kerberos/SASL with auth-int
+    &lt;ul&gt;
+      &lt;li&gt;Builds on auth, also providing message integrity checks of the data going across the wire. You also know that the message you received was not altered.&lt;/li&gt;
+    &lt;/ul&gt;
+  &lt;/li&gt;
+  &lt;li&gt;Kerberos/SASL with auth-conf
+    &lt;ul&gt;
+      &lt;li&gt;Builds on auth-int, also providing confidentiality of the message that was sent to prevent others from reading it (aka wire-encryption).&lt;/li&gt;
+    &lt;/ul&gt;
+  &lt;/li&gt;
+&lt;/ul&gt;
+
+&lt;p&gt;For each test, five iterations were run to obtain a min, max, and median
+time elapsed at each security configuration. After each iteration,
+Hadoop, and Zookeeper processes were restarted, Accumulo tables are
+wiped clean and tables are recreated. In addition, pagecache, dentries
+and inodes are dropped by issuing a ‘3’ command on
+/proc/sys/vm/drop_caches to ensure that the OS is not caching things to disk
+that might affect the benchmark. The following sequence was performed 
+between iterations:&lt;/p&gt;
+
+&lt;ol&gt;
+  &lt;li&gt;Bring down Accumulo&lt;/li&gt;
+  &lt;li&gt;Bring down Zookeeper&lt;/li&gt;
+  &lt;li&gt;Bring down Hadoop&lt;/li&gt;
+  &lt;li&gt;Run sync command&lt;/li&gt;
+  &lt;li&gt;Drop OS cache&lt;/li&gt;
+  &lt;li&gt;Bring up Hadoop&lt;/li&gt;
+  &lt;li&gt;Bring up Zookeeper&lt;/li&gt;
+  &lt;li&gt;Bring up Accumulo&lt;/li&gt;
+  &lt;li&gt;Drop tables&lt;/li&gt;
+  &lt;li&gt;Create tables&lt;/li&gt;
+&lt;/ol&gt;
+
+&lt;p&gt;For each iteration, the results were stored, fed into &lt;a href=&quot;https://nationalsecurityagency.github.io/timely/&quot;&gt;Timely&lt;/a&gt;, and viewed with Grafana.
+Since the runs were executed sequentially, the start epochs for each run did not align.
+To mitigate, the entries for each run were inserted 
+with the same relative epoch for convenient comparison in Grafana.&lt;/p&gt;
+
+&lt;p&gt;The table configurations for Accumulo remain the same throughout the
+different iterations and security levels. The Accumulo site
+configurations differ only due to the different settings for the
+security level configurations.&lt;/p&gt;
+
+&lt;h2 id=&quot;environment&quot;&gt;Environment&lt;/h2&gt;
+
+&lt;p&gt;In order to perform the testing, a small AWS cluster was setup using 14
+hosts on EC2. Two i2.xlarge instances were used as master nodes and eight
+d2.xlarge instances were used for workers. In addition, two c4.4xlarge
+instances were used for ingesters, one m4.2xlarge instance was used for
+Timely, and one m4.xlarge instance was used for Apache Ambari. A logical
+diagram of the setup is depicted below:&lt;/p&gt;
+
+&lt;p&gt;&lt;img src=&quot;/images/blog/201702_security/figure1.png&quot; alt=&quot;&quot; width=&quot;400px&quot; /&gt;&lt;/p&gt;
+
+&lt;p&gt;Figure 1 - Cluster Layout, Roles, and Instance Types on AWS.&lt;/p&gt;
+
+&lt;p&gt;The types of nodes and their function are given below:&lt;/p&gt;
+
+&lt;table id=&quot;instance_types&quot; class=&quot;table&quot;&gt;
+  &lt;thead&gt;
+    &lt;tr&gt;
+      &lt;th style=&quot;text-align: left&quot;&gt;Node Type&lt;/th&gt;
+      &lt;th style=&quot;text-align: left&quot;&gt;AWS EC2 Type&lt;/th&gt;
+      &lt;th style=&quot;text-align: left&quot;&gt;EC2 Type Details&lt;/th&gt;
+      &lt;th style=&quot;text-align: left&quot;&gt;Quantity&lt;/th&gt;
+    &lt;/tr&gt;
+  &lt;/thead&gt;
+  &lt;tbody&gt;
+    &lt;tr&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;Ingest Nodes&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;c4.4xlarge&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;16 core, 30 GB RAM&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;2&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;Worker Node&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;d2.xlarge&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;4 cores, 30.5 GB RAM, 3x2T GB HD&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;8&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;Master Node&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;i2.xlarge&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;4 cores, 30.5 GB RAM, 1x800GB SSD&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;2&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;Admin Node&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;m4.xlarge&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;4 cores, 16 GB RAM&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;1&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;Timely Node&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;m4.2xlarge&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;8 cores, 32 GB RAM&lt;/td&gt;
+      &lt;td style=&quot;text-align: left&quot;&gt;1&lt;/td&gt;
+    &lt;/tr&gt;
+  &lt;/tbody&gt;
+&lt;/table&gt;
+
+&lt;p&gt;Table 1 – AWS Instance Types, Role, Details, and Quantities&lt;/p&gt;
+
+&lt;h2 id=&quot;results&quot;&gt;Results&lt;/h2&gt;
+
+&lt;p&gt;The median, max, and min of the milliseconds elapsed
+time of all iterations for each test is displayed below. The percentage change
+columns compare the Median, Max, and Min respectively from the no
+security level to each security configuration (e.g. no security Median
+vs. auth-int Median, no security Max vs. auth-int Max).&lt;/p&gt;
+
+&lt;table id=&quot;results&quot; class=&quot;table&quot;&gt;
+  &lt;thead&gt;
+    &lt;tr&gt;
+      &lt;th&gt;Security Level&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;Median&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;Standard Deviation&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;Max&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;Min&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;% Change (nosec Median vs. Median)&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;% Change (nosec Max vs. Max)&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;% Change (nosec Min vs. Min)&lt;/th&gt;
+      &lt;th style=&quot;text-align: right&quot;&gt;Delta from Previous Level (Median)&lt;/th&gt;
+    &lt;/tr&gt;
+  &lt;/thead&gt;
+  &lt;tbody&gt;
+    &lt;tr&gt;
+      &lt;td&gt;no security&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;7829394&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;139340&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;8143035&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;7764309&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;0.00%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;0.00%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;0.00%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;0.00%&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td&gt;ssl&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;8292760&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;87012&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;8464060&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;8204955&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;5.92%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;3.94%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;5.68%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;5.92%&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td&gt;auth&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;8859552&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;134109&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9047971&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;8657618&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;13.16%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;11.11%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;11.51%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;6.83%&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td&gt;auth-int&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9500737&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;155968&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9753424&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9282371&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;21.34%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;19.78%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;19.55%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;7.24%&lt;/td&gt;
+    &lt;/tr&gt;
+    &lt;tr&gt;
+      &lt;td&gt;auth-conf&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9479635&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;170823&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9776580&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;9282189&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;21.08%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;20.06%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;19.55%&lt;/td&gt;
+      &lt;td style=&quot;text-align: right&quot;&gt;-0.22%&lt;/td&gt;
+    &lt;/tr&gt;
+  &lt;/tbody&gt;
+&lt;/table&gt;
+
+&lt;p&gt;Table 2 – Summarized Time Elapsed for Each Security Level&lt;/p&gt;
+
+&lt;h2 id=&quot;plots&quot;&gt;Plots&lt;/h2&gt;
+
+&lt;p&gt;Below are some snapshots of *stats.out elements via Grafana that were inserted
+into Timely with the same relative start time.  Each graph represents a field 
+in the output generated by &lt;a href=&quot;https://github.com/apache/accumulo/blob/1.7/test/src/main/java/org/apache/accumulo/test/continuous/ContinuousStatsCollector.java&quot;&gt;ContinuousStatsCollector&lt;/a&gt;&lt;/p&gt;
+
+&lt;h3 id=&quot;tablerecshttpsgithubcomapacheaccumuloblob17coresrcmainjavaorgapacheaccumulocoremasterthrifttableinfojaval73&quot;&gt;&lt;a href=&quot;https://github.com/apache/accumulo/blob/1.7/core/src/main/java/org/apache/accumulo/core/master/thrift/TableInfo.java#L73&quot;&gt;TABLE_RECS&lt;/a&gt;&lt;/h3&gt;
+&lt;p&gt;(Number of records in the continuous ingest table.  Down sample=1m, aggregate=avg)&lt;/p&gt;
+
+&lt;p&gt;&lt;a href=&quot;/images/blog/201702_security/tableRecs.png&quot;&gt;&lt;img src=&quot;/images/blog/201702_security/tableRecs.png&quot; alt=&quot;&quot; width=&quot;800px&quot; /&gt;&lt;/a&gt;&lt;/p&gt;
+
+&lt;h3 id=&quot;totalingesthttpsgithubcomapacheaccumuloblob17coresrcmainjavaorgapacheaccumulocoremasterthrifttableinfojaval77&quot;&gt;&lt;a href=&quot;https://github.com/apache/accumulo/blob/1.7/core/src/main/java/org/apache/accumulo/core/master/thrift/TableInfo.java#L77&quot;&gt;TOTAL_INGEST&lt;/a&gt;&lt;/h3&gt;
+&lt;p&gt;(Ingest rate for Accumulo instance.  Down sample=5m, aggregate=avg)&lt;/p&gt;
+
+&lt;p&gt;&lt;a href=&quot;/images/blog/201702_security/totalIngest.png&quot;&gt;&lt;img src=&quot;/images/blog/201702_security/totalIngest.png&quot; alt=&quot;&quot; width=&quot;800px&quot; /&gt;&lt;/a&gt;&lt;/p&gt;
+
+&lt;h3 id=&quot;avgfilestablethttpsgithubcomapacheaccumuloblob17coresrcmainjavaorgapacheaccumulocoreutilstatjaval63&quot;&gt;&lt;a href=&quot;https://github.com/apache/accumulo/blob/1.7/core/src/main/java/org/apache/accumulo/core/util/Stat.java#L63&quot;&gt;AVG_FILES/TABLET&lt;/a&gt;&lt;/h3&gt;
+&lt;p&gt;(Average number of files per Accumulo tablet.  Down sample=1m, aggregate=avg)&lt;/p&gt;
+
+&lt;p&gt;&lt;a href=&quot;/images/blog/201702_security/avgFilesTab.png&quot;&gt;&lt;img src=&quot;/images/blog/201702_security/avgFilesTab.png&quot; alt=&quot;&quot; width=&quot;800px&quot; /&gt;&lt;/a&gt;&lt;/p&gt;
+
+&lt;h3 id=&quot;accumulofileshttpsgithubcomapacheaccumuloblob17testsrcmainjavaorgapacheaccumulotestcontinuouscontinuousstatscollectorjaval127&quot;&gt;&lt;a href=&quot;https://github.com/apache/accumulo/blob/1.7/test/src/main/java/org/apache/accumulo/test/continuous/ContinuousStatsCollector.java#L127&quot;&gt;ACCUMULO_FILES&lt;/a&gt;&lt;/h3&gt;
+&lt;p&gt;(Total number of files for Accumulo.  Down sample=1m, aggregate=avg)&lt;/p&gt;
+
+&lt;p&gt;&lt;a href=&quot;/images/blog/201702_security/accumuloFiles.png&quot;&gt;&lt;img src=&quot;/images/blog/201702_security/accumuloFiles.png&quot; alt=&quot;&quot; width=&quot;800px&quot; /&gt;&lt;/a&gt;&lt;/p&gt;
+
+&lt;p&gt;As can be seen in the plots above, the different security settings have 
+relatively consistent, discernable median run characteristics.  The big
+dip in each TOTAL_INGEST coincides with a large number of major
+compactions, a rate decrease for TABLE_RECS, and a decrease in 
+AVG_FILES/TABLET.&lt;/p&gt;
+
+&lt;h2 id=&quot;final-thoughts&quot;&gt;Final Thoughts&lt;/h2&gt;
+
+&lt;p&gt;The biggest performance 
+hits to run duration median (compared to default security) were ~21% for 
+auth-int and auth-conf.  Interesting to note that SSL’s median run duration was 
+lower than all SASL configs and that auth-conf’s was lower than auth-int. 
+Initial  speculation for these oddities revolved around the 
+&lt;a href=&quot;https://github.com/m1ch1/mapkeeper/wiki/Thrift-Java-Servers-Compared&quot;&gt;Thrift server&lt;/a&gt; 
+implementations, but the Thrift differences will not explain the auth-conf/int 
+disparity since both utilize TThreadPoolServer.  It was certainly unexpected that the 
+addition of wire encryption would yield a faster median run duration.  This result 
+prompted, as a sanity check, sniffing the net traffic (in a contrived example 
+not during a timed run) in both auth-conf and auth-int to ensure that the message 
+contents were actually obfuscated in auth-conf (they were) and not obfuscated in 
+auth-int (they weren’t).&lt;/p&gt;
+
+&lt;h2 id=&quot;future-work&quot;&gt;Future Work&lt;/h2&gt;
+
+&lt;p&gt;Part 2 of this series will consist of the same continuous ingest loads and 
+configurations with the addition of a query load on the system.&lt;/p&gt;
+
+</description>
+        <pubDate>Thu, 23 Feb 2017 00:00:00 -0500</pubDate>
+        <link>https://accumulo.apache.org/blog/2017/02/23/security-performance-implications.html</link>
+        <guid isPermaLink="true">https://accumulo.apache.org/blog/2017/02/23/security-performance-implications.html</guid>
+        
+        
+        <category>blog</category>
+        
+      </item>
+    
+      <item>
         <title>Running Accumulo on Fedora 25</title>
         <description>&lt;p&gt;Apache Accumulo has been available in &lt;a href=&quot;https://getfedora.org/&quot;&gt;Fedora&lt;/a&gt; since F20. Recently, the Fedora
 packages have been updated to Accumulo version &lt;code class=&quot;highlighter-rouge&quot;&gt;1.6.6&lt;/code&gt; and have made some
@@ -1725,97 +2014,5 @@ HDFS High-Availability instances, forcing NameNode failover.&lt;/p&gt;
         
       </item>
     
-      <item>
-        <title>Apache Accumulo 1.6.4</title>
-        <description>&lt;p&gt;Apache Accumulo 1.6.4 is a maintenance release on the 1.6 version branch.
-This release contains changes from 21 issues, comprised of bug-fixes,
-performance improvements and better test cases. See &lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO/fixforversion/12332840&quot;&gt;JIRA&lt;/a&gt; for a
-complete list.&lt;/p&gt;
-
-&lt;p&gt;Below are resources for this release:&lt;/p&gt;
-
-&lt;ul&gt;
-  &lt;li&gt;&lt;a href=&quot;/1.6/accumulo_user_manual.html&quot;&gt;User Manual&lt;/a&gt;&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;/1.6/apidocs&quot;&gt;Javadocs&lt;/a&gt;&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;/1.6/examples&quot;&gt;Examples&lt;/a&gt;&lt;/li&gt;
-&lt;/ul&gt;
-
-&lt;p&gt;Users of any previous 1.6.x release are strongly encouraged to update as soon as
-possible to benefit from the improvements with very little concern in change
-of underlying functionality. Users of 1.4 or 1.5 that are seeking to upgrade
-to 1.6 should consider 1.6.4 as a starting point.&lt;/p&gt;
-
-&lt;h2 id=&quot;silent-data-loss-via-bulk-imported-files&quot;&gt;Silent data-loss via bulk imported files&lt;/h2&gt;
-
-&lt;p&gt;A user recently reported that a simple bulk-import application would occasionally
-lose some records. Through investigation, it was found that when bulk imports into
-a table failed the initial assignment, the logic that automatically retries the
-imports was incorrectly choosing the tablets to import the files into. &lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3967&quot;&gt;ACCUMULO-3967&lt;/a&gt;
-contains more information on the cause and identification of the bug. The data-loss
-condition would only affect entire files. If records from a file exist in Accumulo,
-it is still guaranteed that all records within that imported file were successful.&lt;/p&gt;
-
-&lt;p&gt;As such, users who have bulk import applications using previous versions of Accumulo
-should verify that all of their data was correctly ingested into Accumulo and
-immediately update to Accumulo 1.6.4.&lt;/p&gt;
-
-&lt;h2 id=&quot;other-bug-fixes&quot;&gt;Other bug fixes&lt;/h2&gt;
-
-&lt;ul&gt;
-  &lt;li&gt;&lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3979&quot;&gt;ACCUMULO-3979&lt;/a&gt; Fixed an issue where the BulkImporter failed
-with an error message “QUERY_METADATA already started”.&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3965&quot;&gt;ACCUMULO-3965&lt;/a&gt; The &lt;code class=&quot;highlighter-rouge&quot;&gt;listscans&lt;/code&gt; shell command did not contain
-the &lt;code class=&quot;highlighter-rouge&quot;&gt;scanId&lt;/code&gt; attribute for currently running scans.&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3946&quot;&gt;ACCUMULO-3946&lt;/a&gt; Verified that all user-facing operations contained
-appropriate audit messages.&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3977&quot;&gt;ACCUMULO-3977&lt;/a&gt; Isolated scans with Iterators in use incorrectly
-fail around invocation of &lt;code class=&quot;highlighter-rouge&quot;&gt;deepCopy&lt;/code&gt;.&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3905&quot;&gt;ACCUMULO-3905&lt;/a&gt; RowDeletingIterator functions incorrectly when
-columns are provided by the client. This restores intended functionality without
-the need for a &lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-1801?focusedCommentId=13970204&amp;amp;page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13970204&quot;&gt;workaround&lt;/a&gt;.&lt;/li&gt;
-  &lt;li&gt;&lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3959&quot;&gt;ACCUMULO-3959&lt;/a&gt; &lt;a href=&quot;https://issues.apache.org/jira/browse/ACCUMULO-3934&quot;&gt;ACCUMULO-3934&lt;/a&gt; Multiple documentation
-improvements to &lt;code class=&quot;highlighter-rouge&quot;&gt;BatchScanner&lt;/code&gt;.&lt;/li&gt;
-&lt;/ul&gt;
-
-&lt;h2 id=&quot;testing&quot;&gt;Testing&lt;/h2&gt;
-
-&lt;p&gt;Each unit and functional test only runs on a single node, while the RandomWalk
-and Continuous Ingest tests run on any number of nodes. &lt;em&gt;Agitation&lt;/em&gt; refers to
-randomly restarting Accumulo processes and Hadoop Datanode processes, and, in
-HDFS High-Availability instances, forcing NameNode failover.&lt;/p&gt;
-
-&lt;table id=&quot;release_notes_testing&quot; class=&quot;table&quot;&gt;
-  &lt;thead&gt;
-    &lt;tr&gt;
-      &lt;th&gt;OS&lt;/th&gt;
-      &lt;th&gt;Hadoop&lt;/th&gt;
-      &lt;th&gt;Nodes&lt;/th&gt;
-      &lt;th&gt;ZooKeeper&lt;/th&gt;
-      &lt;th&gt;HDFS HA&lt;/th&gt;
-      &lt;th&gt;Tests&lt;/th&gt;
-    &lt;/tr&gt;
-  &lt;/thead&gt;
-  &lt;tbody&gt;
-    &lt;tr&gt;
-      &lt;td&gt;Amazon Linux 2014.09&lt;/td&gt;
-      &lt;td&gt;2.6.0&lt;/td&gt;
-      &lt;td&gt;20&lt;/td&gt;
-      &lt;td&gt;3.4.5&lt;/td&gt;
-      &lt;td&gt;No&lt;/td&gt;
-      &lt;td&gt;ContinuousIngest w/ verification w/ and w/o agitation (37B entries)&lt;/td&gt;
-    &lt;/tr&gt;
-  &lt;/tbody&gt;
-&lt;/table&gt;
-
-</description>
-        <pubDate>Sat, 03 Oct 2015 00:00:00 -0400</pubDate>
-        <link>https://accumulo.apache.org/release/accumulo-1.6.4/</link>
-        <guid isPermaLink="true">https://accumulo.apache.org/release/accumulo-1.6.4/</guid>
-        
-        
-        <category>release</category>
-        
-      </item>
-    
   </channel>
 </rss>

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/images/blog/201702_security/accumuloFiles.png
----------------------------------------------------------------------
diff --git a/images/blog/201702_security/accumuloFiles.png b/images/blog/201702_security/accumuloFiles.png
new file mode 100644
index 0000000..4c23277
Binary files /dev/null and b/images/blog/201702_security/accumuloFiles.png differ

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/images/blog/201702_security/avgFilesTab.png
----------------------------------------------------------------------
diff --git a/images/blog/201702_security/avgFilesTab.png b/images/blog/201702_security/avgFilesTab.png
new file mode 100644
index 0000000..8251bca
Binary files /dev/null and b/images/blog/201702_security/avgFilesTab.png differ

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/images/blog/201702_security/figure1.png
----------------------------------------------------------------------
diff --git a/images/blog/201702_security/figure1.png b/images/blog/201702_security/figure1.png
new file mode 100644
index 0000000..f28715c
Binary files /dev/null and b/images/blog/201702_security/figure1.png differ

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/images/blog/201702_security/tableRecs.png
----------------------------------------------------------------------
diff --git a/images/blog/201702_security/tableRecs.png b/images/blog/201702_security/tableRecs.png
new file mode 100644
index 0000000..b7db4ee
Binary files /dev/null and b/images/blog/201702_security/tableRecs.png differ

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/images/blog/201702_security/tableRecsInMem.png
----------------------------------------------------------------------
diff --git a/images/blog/201702_security/tableRecsInMem.png b/images/blog/201702_security/tableRecsInMem.png
new file mode 100644
index 0000000..b480372
Binary files /dev/null and b/images/blog/201702_security/tableRecsInMem.png differ

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/images/blog/201702_security/totalIngest.png
----------------------------------------------------------------------
diff --git a/images/blog/201702_security/totalIngest.png b/images/blog/201702_security/totalIngest.png
new file mode 100644
index 0000000..3bc3626
Binary files /dev/null and b/images/blog/201702_security/totalIngest.png differ

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/index.html
----------------------------------------------------------------------
diff --git a/index.html b/index.html
index cef80f8..0ff029d 100644
--- a/index.html
+++ b/index.html
@@ -164,29 +164,29 @@
         
         <div class="row latest-news-item">
           <div class="col-sm-12" style="margin-bottom: 5px">
-           <span style="font-size: 12px; margin-right: 5px;">Dec 2016</span>
-           <a href="/blog/2016/12/19/running-on-fedora-25.html">Running Accumulo on Fedora 25</a>
+           <span style="font-size: 12px; margin-right: 5px;">Feb 2017</span>
+           <a href="/blog/2017/02/23/security-performance-implications.html">Security Performance Implications</a>
           </div>
         </div>
         
         <div class="row latest-news-item">
           <div class="col-sm-12" style="margin-bottom: 5px">
-           <span style="font-size: 12px; margin-right: 5px;">Nov 2016</span>
-           <a href="/blog/2016/11/16/simpler-scripts-and-config.html">Simpler scripts and configuration coming in Accumulo 2.0.0</a>
+           <span style="font-size: 12px; margin-right: 5px;">Dec 2016</span>
+           <a href="/blog/2016/12/19/running-on-fedora-25.html">Running Accumulo on Fedora 25</a>
           </div>
         </div>
         
         <div class="row latest-news-item">
           <div class="col-sm-12" style="margin-bottom: 5px">
            <span style="font-size: 12px; margin-right: 5px;">Nov 2016</span>
-           <a href="/blog/2016/11/02/durability-performance.html">Durability Performance Implications</a>
+           <a href="/blog/2016/11/16/simpler-scripts-and-config.html">Simpler scripts and configuration coming in Accumulo 2.0.0</a>
           </div>
         </div>
         
         <div class="row latest-news-item">
           <div class="col-sm-12" style="margin-bottom: 5px">
-           <span style="font-size: 12px; margin-right: 5px;">Sep 2016</span>
-           <a href="/release/accumulo-1.6.6/">Apache Accumulo 1.6.6</a>
+           <span style="font-size: 12px; margin-right: 5px;">Nov 2016</span>
+           <a href="/blog/2016/11/02/durability-performance.html">Durability Performance Implications</a>
           </div>
         </div>
         

http://git-wip-us.apache.org/repos/asf/accumulo-website/blob/803c95d0/news/index.html
----------------------------------------------------------------------
diff --git a/news/index.html b/news/index.html
index bd07ba2..2ffbe5c 100644
--- a/news/index.html
+++ b/news/index.html
@@ -155,6 +155,13 @@
 
   
   
+  <div class="row" style="margin-top: 15px">
+    <div class="col-md-1">Feb 23</div>
+    <div class="col-md-10"><a href="/blog/2017/02/23/security-performance-implications.html">Security Performance Implications</a></div>
+  </div>
+
+  
+  
     
     <hr />
     <h3>2016</h3>


Mime
View raw message