HP XC System Software Administration Guide Version 2.1

6
Monitoring the System
System monitoring can identify situations th at can become problems later. This chapter
discusses the following topics:
• An overview of the m o nitoring to ols (Sec
tion 6.2)
• How to observe system environmental data (Section 6.3)
• How to view system statistics (Section 6.4)
• How to customize Nagios metrics gathe
ring (Section 6.5)
• Organization of t he logging of node events and how to view those events (Section 6.6)
6.1 Monitoring Strategy
The H P XC system monitoring strateg y is built on the Nagios system, the Supermon monitors,
and the syslog-ng logging system. I t is designed in a tiered setup to allow for vary ing levels of
granularity in logging and monito ring information.
For monitoring pu rpo ses, the nodes can be considered as a hierarchy, where local nodes talk
to manage men t hubs (regional nodes), which in turn talk to a central man agem ent console
(global node). Figure 6-1 shows this structure.
Figure 6-1: Tiered Structure for Node Events
Local
Node
Local
Node
Local
Node
Local
Node
Local
Node
Local
Node
Global
Node
Regional
Node
Regional
Node
6.2 Monitoring Tools
The HP XC System Software includes t he shownode metrics command in addition to
standard Linux monitoring commands. The HP XC system also includes the Nagios Web-based
utility fo r s ystem monitoring.
6.2.1 Commands for Monitoring Node Status
The shownode metrics comm and, which can be issued from any node in the HP XC
system, provides the ability t o mon ito r the status of all the nodes in t he system.
The following arguments to the shownode metrics co mmand monitor the node status:
• shownode metrics cpus
• shownode metrics cputotals
• shownode metrics load
• shownode metrics mem
Monitoring the System 6-1