|
|
@@ -0,0 +1,429 @@
|
|
|
+<?xml version="1.0" encoding="utf-8"?>
|
|
|
+<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN"
|
|
|
+"http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd">
|
|
|
+<book lang="en_US" xml:base="../">
|
|
|
+ <bookinfo>
|
|
|
+ <title>HPCC Monitoring and Reporting (Technical Preview)</title>
|
|
|
+
|
|
|
+ <mediaobject>
|
|
|
+ <imageobject>
|
|
|
+ <imagedata fileref="images/redswooshWithLogo3.jpg" />
|
|
|
+ </imageobject>
|
|
|
+ </mediaobject>
|
|
|
+
|
|
|
+ <author>
|
|
|
+ <surname>Boca Raton Documentation Team</surname>
|
|
|
+ </author>
|
|
|
+
|
|
|
+ <legalnotice>
|
|
|
+ <para>We welcome your comments and feedback about this document via
|
|
|
+ email to <email>docfeedback@hpccsystems.com</email> Please include
|
|
|
+ <emphasis role="bold">Documentation Feedback</emphasis> in the subject
|
|
|
+ line and reference the document name, page numbers, and current Version
|
|
|
+ Number in the text of the message.</para>
|
|
|
+
|
|
|
+ <para>LexisNexis and the Knowledge Burst logo are registered trademarks
|
|
|
+ of Reed Elsevier Properties Inc., used under license. HPCC Systems is a
|
|
|
+ registered trademark of LexisNexis Risk Data Management Inc.</para>
|
|
|
+
|
|
|
+ <para>Other products, logos, and services may be trademarks or
|
|
|
+ registered trademarks of their respective companies. All names and
|
|
|
+ example data used in this manual are fictitious. Any similarity to
|
|
|
+ actual persons, living or dead, is purely coincidental.</para>
|
|
|
+
|
|
|
+ <para></para>
|
|
|
+ </legalnotice>
|
|
|
+
|
|
|
+ <xi:include href="common/Version.xml" xpointer="FooterInfo"
|
|
|
+ xmlns:xi="http://www.w3.org/2001/XInclude" />
|
|
|
+
|
|
|
+ <xi:include href="common/Version.xml" xpointer="DateVer"
|
|
|
+ xmlns:xi="http://www.w3.org/2001/XInclude" />
|
|
|
+
|
|
|
+ <corpname>HPCC Systems</corpname>
|
|
|
+
|
|
|
+ <xi:include href="common/Version.xml" xpointer="Copyright"
|
|
|
+ xmlns:xi="http://www.w3.org/2001/XInclude" />
|
|
|
+
|
|
|
+ <mediaobject role="logo">
|
|
|
+ <imageobject>
|
|
|
+ <imagedata fileref="images/LN_Rightjustified.jpg" />
|
|
|
+ </imageobject>
|
|
|
+ </mediaobject>
|
|
|
+ </bookinfo>
|
|
|
+
|
|
|
+ <chapter id="GangliaIntroduction">
|
|
|
+ <title>Introduction</title>
|
|
|
+
|
|
|
+ <para>The HPCC Systems platform supports a graphical monitoring and
|
|
|
+ reporting component. With the the graphical monitoring component you can:
|
|
|
+ <itemizedlist>
|
|
|
+ <listitem>
|
|
|
+ <para>See system information at a glance</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>View a grid of Roxie clusters</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Examine Roxie metrics</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Keep a historical record of metrics</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Drill down to individual server metrics</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Quickly detect troubled nodes</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>More applications, such as better informed resource planning
|
|
|
+ and allocation</para>
|
|
|
+ </listitem>
|
|
|
+ </itemizedlist></para>
|
|
|
+
|
|
|
+ <para>The HPCC monitoring component leverages Ganglia, an open-source,
|
|
|
+ scalable, distributed monitoring system to display system information in a
|
|
|
+ graphical manner.</para>
|
|
|
+
|
|
|
+ <!--***NOTE: At some point this next bit will need to get moved into the next chapter/section-->
|
|
|
+
|
|
|
+ <sect1 id="HPCC_Viewer">
|
|
|
+ <title>The HPCC Ganglia Viewer</title>
|
|
|
+
|
|
|
+ <para>A Ganglia viewer comes preinstalled and configured in the 4.2.x
|
|
|
+ (or later) HPCC Virtual Machine. The monitoring provided with the
|
|
|
+ Virtual Machine is set up to monitor Roxie instances on the network.
|
|
|
+ This document introduces the monitoring and describes how to get it
|
|
|
+ working on your system. <figure>
|
|
|
+ <title>HPCC Monitoring</title>
|
|
|
+
|
|
|
+ <mediaobject>
|
|
|
+ <imageobject>
|
|
|
+ <imagedata fileref="images/GAN001.jpg" vendor="Ganglia" />
|
|
|
+ </imageobject>
|
|
|
+ </mediaobject>
|
|
|
+ </figure></para>
|
|
|
+
|
|
|
+ <para>The above image is an overview summary of all the monitored Roxie
|
|
|
+ nodes in the cluster named VM Demo.</para>
|
|
|
+ </sect1>
|
|
|
+ </chapter>
|
|
|
+
|
|
|
+ <chapter id="Ganglya_Overview">
|
|
|
+ <title>Ganglia</title>
|
|
|
+
|
|
|
+ <para>The HPCC Monitoring component leverages Ganglia, an open-source,
|
|
|
+ scalable, distributed monitoring system, to produce a graphical view of a
|
|
|
+ Roxie cluster's servers. Ganglia leverages widely accepted technologies
|
|
|
+ for data representation. It provides near real-time monitoring and
|
|
|
+ visualizations for performance metrics. If your enterprise already has a
|
|
|
+ Ganglia monitoring server, you can easily add Roxie clusters to its
|
|
|
+ monitoring.</para>
|
|
|
+
|
|
|
+ <sect1 id="Ganglia_Overview">
|
|
|
+ <title>Ganglia Overview</title>
|
|
|
+
|
|
|
+ <para>Ganglia Monitoring has two primary components: the viewer and the
|
|
|
+ Ganglia Monitoring Daemon (gmond). Installation and configuration will
|
|
|
+ vary depending on your system.</para>
|
|
|
+
|
|
|
+ <para>On an <emphasis role="bold">RPM-based system</emphasis>, install
|
|
|
+ the <emphasis>ganglia-gmond-modules-python</emphasis> RPM running a
|
|
|
+ command such as:</para>
|
|
|
+
|
|
|
+ <programlisting>sudo rpm -i ganglia-gmond-modules-python-3.4.0-1.x86_64.rpm</programlisting>
|
|
|
+
|
|
|
+ <para>On a <emphasis role="bold">Debian-based system</emphasis>, install
|
|
|
+ the <emphasis>ganglia-monitor</emphasis> running a command such
|
|
|
+ as:</para>
|
|
|
+
|
|
|
+ <programlisting>sudo apt-get install ganglia-monitor</programlisting>
|
|
|
+
|
|
|
+ <para>The specific steps required to install and configure Ganglia are
|
|
|
+ covered in the Ganglia wiki:</para>
|
|
|
+
|
|
|
+ <para><ulink
|
|
|
+ url="http://sourceforge.net/apps/trac/ganglia/wiki/ganglia_gmond_python_modules">http://sourceforge.net/apps/trac/ganglia/wiki/ganglia_gmond_python_modules</ulink></para>
|
|
|
+
|
|
|
+ <sect2 id="Ganglia_Viewer">
|
|
|
+ <title>The Viewer</title>
|
|
|
+
|
|
|
+ <para>If you already have a Ganglia monitoring server running in your
|
|
|
+ network, the viewer component may already be in place. If you do not
|
|
|
+ have Ganglia then you will need to install and configure the
|
|
|
+ viewer.</para>
|
|
|
+
|
|
|
+ <para>Refer to the <ulink
|
|
|
+ url="https://github.com/hpcc-systems/ganglia-monitoring/tree/master/vm_precise">https://github.com/hpcc-systems/ganglia-monitoring/tree/master/vm_precise</ulink>
|
|
|
+ directory. There you will see the resources used to configure the
|
|
|
+ Ganglia for the virtual machine and can use them as the examples to
|
|
|
+ configure it for your enterprise.</para>
|
|
|
+
|
|
|
+ <para>The script, <emphasis>install_graphs_helper.sh</emphasis>
|
|
|
+ available from the github link above and also provided with the
|
|
|
+ virtual machine, is what is used to embed the viewer component. Using
|
|
|
+ this script as a basis, you can then similarly configure and deploy
|
|
|
+ the viewer component for your system.</para>
|
|
|
+ </sect2>
|
|
|
+
|
|
|
+ <sect2 id="Ganglia_Monitoring_Daemon">
|
|
|
+ <title>Ganglia Monitoring Daemon</title>
|
|
|
+
|
|
|
+ <para>The monitoring daemon is required on the Roxie nodes you wish to
|
|
|
+ monitor. Install the <emphasis>gmond</emphasis> daemon on the nodes
|
|
|
+ you wish to monitor. Installation and configuration are described
|
|
|
+ in;</para>
|
|
|
+
|
|
|
+ <para><ulink
|
|
|
+ url="http://sourceforge.net/apps/trac/ganglia/wiki/ganglia_gmond_python_modules">http://sourceforge.net/apps/trac/ganglia/wiki/ganglia_gmond_python_modules</ulink></para>
|
|
|
+
|
|
|
+ <para>If you have a Ganglia monitoring server running in your
|
|
|
+ environment, you already have the required components and
|
|
|
+ prerequisites. Verify that you have
|
|
|
+ <emphasis>/etc/ganglia/conf.d</emphasis> and
|
|
|
+ <emphasis>/etc/ganglia/.pyconf</emphasis> files in place and then add
|
|
|
+ the Roxie nodes you wish to monitor. You can do that by installing the
|
|
|
+ Ganglia components and HPCC Monitoring components on to each Roxie
|
|
|
+ node.</para>
|
|
|
+
|
|
|
+ <para>If you do not have Ganglia, or want to install it, read the
|
|
|
+ Ganglia documentation provided at the above link, and install it and
|
|
|
+ any system dependencies. You will then need to download and install
|
|
|
+ the HPCC Monitoring component.</para>
|
|
|
+
|
|
|
+ <sect3 id="Installing-HPCCGanglia" role="brk">
|
|
|
+ <title>Installing the HPCC Monitoring component</title>
|
|
|
+
|
|
|
+ <para>The HPCC Monitoring component is available for download. The
|
|
|
+ HPCC Monitoring components leverage the Ganglia monitoring tools,
|
|
|
+ and would only be needed if you do not already have Ganglia
|
|
|
+ monitoring components on your system.</para>
|
|
|
+
|
|
|
+ <para>To get the HPCC Monitoring components, find the appropriate
|
|
|
+ package for your system.</para>
|
|
|
+
|
|
|
+ <para>Packages are available for download from the HPCC Systems
|
|
|
+ site:</para>
|
|
|
+
|
|
|
+ <para><ulink
|
|
|
+ url="http://hpccsystems.com/download">hpccsystems.com/download</ulink></para>
|
|
|
+
|
|
|
+ <para>or</para>
|
|
|
+
|
|
|
+ <para><ulink
|
|
|
+ url="http://hpccsystems.com/download/free-community-edition/all">http://hpccsystems.com/download/free-community-edition/all</ulink></para>
|
|
|
+
|
|
|
+ <para>Find and install the appropriate package for your
|
|
|
+ system.</para>
|
|
|
+
|
|
|
+ <para>For example, if you have a CentOS 6.x system, get the RPM
|
|
|
+ package.</para>
|
|
|
+
|
|
|
+ <programlisting>hpccsystems-ganglia-monitoring-4.2.0-rc1.el6.x86_64.rpm</programlisting>
|
|
|
+
|
|
|
+ <para>Install the monitoring package on the system that you want to
|
|
|
+ monitor. Optionally, you can look at that installation package
|
|
|
+ provided and use that as a guide to implement your own customized
|
|
|
+ monitoring components.</para>
|
|
|
+ </sect3>
|
|
|
+ </sect2>
|
|
|
+ </sect1>
|
|
|
+
|
|
|
+ <sect1 id="VirtuaLMetrics">
|
|
|
+ <title>Metrics in the Virtual Machine</title>
|
|
|
+
|
|
|
+ <para>An easy way to understand how the metrics work and how to
|
|
|
+ implement them on a larger system, is to examine the metrics in
|
|
|
+ action.</para>
|
|
|
+
|
|
|
+ <para>Ganglia integration is built into the current HPCC Virtual Machine
|
|
|
+ images. Download and start up a virtual image and look at how the
|
|
|
+ monitoring component works.</para>
|
|
|
+
|
|
|
+ <para>This allows you:</para>
|
|
|
+
|
|
|
+ <itemizedlist>
|
|
|
+ <listitem>
|
|
|
+ <para>A preview of the metrics</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>A quickstart</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>A guide for set up</para>
|
|
|
+ </listitem>
|
|
|
+ </itemizedlist>
|
|
|
+
|
|
|
+ <para>Evaluate the value of the content and decide what aspects of
|
|
|
+ measurement are relevant to your needs.</para>
|
|
|
+
|
|
|
+ <sect2>
|
|
|
+ <title id="get_hpcc">Get the latest HPCC Virtual Image File</title>
|
|
|
+
|
|
|
+ <para>The complete details for installing and running HPCC in a
|
|
|
+ virtual machine are available in the document: <emphasis
|
|
|
+ role="bold">Running HPCC in a Virtual Machine</emphasis>, available
|
|
|
+ from <ulink
|
|
|
+ url="hpccsystems.com/download/docs">hpccsystems.com/download/docs</ulink>
|
|
|
+ .</para>
|
|
|
+
|
|
|
+ <para>The following steps are a quick summary, assuming you have some
|
|
|
+ familiarity with running virtual machines.</para>
|
|
|
+
|
|
|
+ <para><orderedlist>
|
|
|
+ <listitem>
|
|
|
+ <para>Download the latest HPCC Virtual Machine image file
|
|
|
+ from:</para>
|
|
|
+
|
|
|
+ <para><ulink
|
|
|
+ url="http://HPCCsystems.com/download/hpcc-vm-image">http://hpccsystems.com/download/hpcc-vm-image</ulink></para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Save the file to a folder on your machine.</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Open your virtualization software, import the virtual
|
|
|
+ machine and start it.</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <?dbfo keep-together="always"?>
|
|
|
+
|
|
|
+ <para>Once the VM initialization completes, you will see a
|
|
|
+ window similar to the following:</para>
|
|
|
+
|
|
|
+ <figure id="welcometovm">
|
|
|
+ <title xreflabel="welc">VM Welcome Screen</title>
|
|
|
+
|
|
|
+ <mediaobject>
|
|
|
+ <imageobject>
|
|
|
+ <imagedata fileref="images/GA-vm01.jpg"
|
|
|
+ vendor="VM_welcome" />
|
|
|
+ </imageobject>
|
|
|
+ </mediaobject>
|
|
|
+ </figure>
|
|
|
+
|
|
|
+ <para><informaltable colsep="1" frame="all" rowsep="1">
|
|
|
+ <?dbfo keep-together="always"?>
|
|
|
+
|
|
|
+ <tgroup cols="2">
|
|
|
+ <colspec colwidth="49.50pt" />
|
|
|
+
|
|
|
+ <colspec />
|
|
|
+
|
|
|
+ <tbody>
|
|
|
+ <row>
|
|
|
+ <entry><inlinegraphic
|
|
|
+ fileref="images/caution.png" /></entry>
|
|
|
+
|
|
|
+ <entry>Your virtual IP address could be different from
|
|
|
+ the ones provided in the example images. Please use
|
|
|
+ the IP address provided by <emphasis
|
|
|
+ role="bold">your</emphasis> installation.</entry>
|
|
|
+ </row>
|
|
|
+ </tbody>
|
|
|
+ </tgroup>
|
|
|
+ </informaltable></para>
|
|
|
+
|
|
|
+ <para>Note the IP Address of your VM Instance.</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>In your browser, enter the URL displayed (circled in red
|
|
|
+ above) in the previous image (without the :8010) instead enter
|
|
|
+ the <emphasis>IP Address</emphasis>/ganglia.</para>
|
|
|
+
|
|
|
+ <para>For example,
|
|
|
+ <emphasis>http://nnn.nnn.nnn.nnn/ganglia</emphasis>, where
|
|
|
+ nnn.nnn.nnn.nnn is your Virtual Machine's IP address displayed
|
|
|
+ at the VM welcome screen.</para>
|
|
|
+ </listitem>
|
|
|
+ </orderedlist></para>
|
|
|
+
|
|
|
+ <para>We encourage experienced users to use SSH and log into the VM
|
|
|
+ and further examine the configuration of a 1-node monitoring
|
|
|
+ solution.</para>
|
|
|
+
|
|
|
+ <sect3 id="ViewTheMetrics">
|
|
|
+ <title>Viewing the Metrics</title>
|
|
|
+
|
|
|
+ <para>To view the metrics page, go to the following page in your
|
|
|
+ browser.<programlisting> <emphasis>http://nnn.nnn.nnn.nnn/ganglia</emphasis></programlisting></para>
|
|
|
+
|
|
|
+ <para>Where the <emphasis>nnn.nnn.nnn.nnn</emphasis> is your ESP
|
|
|
+ server running ECL Watch.</para>
|
|
|
+ </sect3>
|
|
|
+ </sect2>
|
|
|
+ </sect1>
|
|
|
+
|
|
|
+ <sect1 id="GangliaIntegration">
|
|
|
+ <title>Ganglia Integration with HPCC</title>
|
|
|
+
|
|
|
+ <para>The Roxie nodes are able to report metrics to Ganglia when the
|
|
|
+ nodes have Ganglia monitoring and the associated dependencies
|
|
|
+ installed.</para>
|
|
|
+
|
|
|
+ <para>Review the Ganglia wiki: <ulink
|
|
|
+ url="http://sourceforge.net/apps/trac/ganglia/wiki/ganglia_gmond_python_modules">http://sourceforge.net/apps/trac/ganglia/wiki/ganglia_gmond_python_modules</ulink>
|
|
|
+ to understand the requirements.</para>
|
|
|
+
|
|
|
+ <orderedlist>
|
|
|
+ <listitem>
|
|
|
+ <para>Install the Ganglia components on every node.</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Configure the Ganglia as appropriate for your system</para>
|
|
|
+
|
|
|
+ <para>The Ganglia configuration files can be typically found in the
|
|
|
+ <emphasis>/etc/ganglia/</emphasis> directory.</para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Install the HPCC Systems monitoring component on every node.
|
|
|
+ </para>
|
|
|
+ </listitem>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Deploy the monitoring daemon (gmond) and the HPCC Monitoring
|
|
|
+ package to each of the nodes you wish to monitor.</para>
|
|
|
+ </listitem>
|
|
|
+ </orderedlist>
|
|
|
+
|
|
|
+ <para>The VM graphs can be used to monitor Roxie clusters. You can add
|
|
|
+ more Roxie nodes installed anywhere on the same network utilizing
|
|
|
+ multi-cast.</para>
|
|
|
+
|
|
|
+ <para>To add a new Roxie node, install the HPCC Monitoring package on to
|
|
|
+ each Roxie node to monitor. In most basic configurations you may need to
|
|
|
+ add the node(s) IP address(es) to the
|
|
|
+ <emphasis>/etc/ganglia/gmetad.conf</emphasis> file. As long as the new
|
|
|
+ Roxie node can communicate with (for example ping) the Monitoring
|
|
|
+ component host, the graphs for that will automatically be added to the
|
|
|
+ graph display.</para>
|
|
|
+
|
|
|
+ <para><variablelist>
|
|
|
+ <varlistentry>
|
|
|
+ <term>NOTE:</term>
|
|
|
+
|
|
|
+ <listitem>
|
|
|
+ <para>Some of the graphs take some time to populate with data.
|
|
|
+ These graphs may appear blank or empty at first, but will render
|
|
|
+ properly as more data accumulates to populate the graph.</para>
|
|
|
+ </listitem>
|
|
|
+ </varlistentry>
|
|
|
+ </variablelist></para>
|
|
|
+ </sect1>
|
|
|
+ </chapter>
|
|
|
+</book>
|
g-pan
