ProCKSI

h1. Monitoring of Cluster Resources: Ganglia

_Ganglia is a scalable distributed monitoring system for high-performance computing systems such as clusters and Grids._

* Download the latest release of the _Ganglia Monitoring Core_ from "http://ganglia.sourceforge.net":http://ganglia.sourceforge.net/.

* Install Ganglia into _/usr/local/ganglia_, its web frontend into _/usr/local/ganglia/html/', and its databases into _/usr/local/ganglia/rrds/'.

* Install the _Ganglia Monitoring Daemon_ (gmond) on each node, and the _Ganglia Meta Daemon_ (gmetad) on the the head node.

h2. Ganglia Monitoring Daemon

* Configure, build and install Ganglia on each slave node (only with _gmond_):

<pre>

  ./configure --prefix=/usr/local

</pre>

  and on the master node (with _gmond_ and _gmetad_):

<pre>

  ./configure --prefix=/usr/local --with-gmeta

</pre>

* Initialise the configuration file for _gmond_:

<pre>

  gmond --default >> /etc/gmond.conf

</pre>

* Configure the _Ganglia Monitoring Daemon_ in _/etc/gmond.conf_:

** Set the name of the cluster: 

<pre>

  cluster {

    name = "ProCKSI"

  }

</pre>

** Set the IP address and port for multicast data exchange:

<pre>

  udp_send_channel {

    mcast_join = 239.2.11.71

    port = 8649

  }

  udp_recv_channel {

    mcast_join = 239.2.11.71

    port = 8649

    bind = 239.2.11.71

  }

</pre>

* Copy start-up script for _gmond_:

<pre>

  cp ./gmond/gmond.init /etc/init.d/gmond

</pre>

* Add additional route for correct data exchange via multicast using the _internal_ interface (_eth0_). Modify _/etc/inid.d/gmond_:

<pre>

   #Add multicast route to internal interface

   /sbin/route add -host 239.2.11.71 dev eth0

   daemon $GMOND

</pre>

<pre>

   #Remove multicast route to internal interface

   /sbin/route delete -host 239.2.11.71 dev eth0

   killproc gmond

</pre>

* Make the Ganglia Monitoring Daemon start at bootup.

<pre>

   /sbin/chkconfig  gmond  on

</pre>

* Start the Ganglia Monitoring Daemon:

<pre>

   /sbin/service  gmond  start

</pre>

h2. Ganglia Meta Daemon

* Install and configure the _Ganglia Meta Daeomn_ (gmetad) on the master node.

* Make the Ganglia Meta Daemon start at bootup.

<pre>

   /sbin/chkconfig  --add gmetad

   /sbin/chkconfig  gmetad  on

</pre>

* Start the Ganglia Meta Daemon:

<pre>

   /sbin/service  gmetad  start

</pre>

* If the pie chart diagrams do not show up, you have to install the _php-gd_ packages.

h2. Further Customisation

In order to display more fine-grained time intervals, edit the following files in _/usr/local/ganglia/html/_:

* *header.php*

<pre>

  if (!$physical) {

   $context_ranges[]="10 minutes";

   $context_ranges[]="20 minutes";

   $context_ranges[]="30 minutes";

   $context_ranges[]="1 hour";

   $context_ranges[]="2 hours";

   $context_ranges[]="4 hours";

   $context_ranges[]="8 hours";

   $context_ranges[]="12 hours";

   $context_ranges[]="1 day";

   $context_ranges[]="2 days";

   $context_ranges[]="week";

   $context_ranges[]="month";

   $context_ranges[]="year";

</pre>

* *get_context.php*

<pre>

  switch ($range) {

   case "10 minutes":   $start = -600; break;

   case "20 minutes":   $start = -1200; break;

   case "30 minutes":   $start = -1800; break;

   case "1 hour":       $start = -3600; break;

   case "2 hours":      $start = -7200; break;

   case "4 hours":      $start = -14400; break;

   case "8 hours":      $start = -28800; break;

   case "12 hours":     $start = -43200; break;

   case "1 day":        $start = -86400; break;

   case "2 days":       $start = -172800; break;

   case "week":         $start = -604800; break;

   case "month":        $start = -2419200; break;

   case "year":         $start = -31449600; break;

</pre>