The basic Nagios set-up consists of a central server that polls all the Hadoop daemons - Hadoop master and slave nodes, the Templeton server, Zookeeper nodes, and HBase servers - in your installation.
Nagios checks these services for one of the following states:
OK
Warning
Critical
Nagios writes its output to a status log,
/var/nagios/status.dat
. The alert information from that file is
periodically collected and displayed in the HDP Monitoring Dashboard, For
more details on Nagios architecture, see the Nagios Overview at
the nagios.org
web site.
Out of the box Hortonworks Data Platform (HDP) provides a set of Nagios plugins specially designed for monitoring important aspects of your Hadoop cluster.