Example 1: Solving A Cpu Performance Problem; Event And Health Monitoring - HP BL685c - ProLiant - 4 GB RAM Introduction Manual

Example 1: Solving a CPU performance problem

Keeping hundreds or thousands of server blades running on an ongoing basis would be difficult
without automation tools. These automation tools can manage large numbers of servers from a single
console. Insight Control and Server Automation reduce the effort required to do this. In general,
Insight Control monitors, manages, and controls the server hardware, and Server Automation monitors
the software configuration compliance, yielding a complete picture.
In this example, the system administrator is using HP SIM, the foundation of Insight Control, for health
monitoring and fault management. A pager alert is received from HP SIM indicating that one of the
servers has gone over its CPU threshold. The system administrator uses the real time graphing
capability of PMP to confirm that this system is consistently running at a higher CPU level than it
should be for that time of day. A quick look at the Compliance Dashboard in Server Automation
shows that Oracle configuration for this server is out of compliance. To reinstate the original
configuration values, the system administrator clicks Remediate. PMP confirms that the system is
returning to its expected CPU usage levels.

Event and health monitoring

In this data center, the system administrator routinely monitors the infrastructure using HP SIM. HP SIM
polls each managed server periodically to retrieve a consolidated health status value that reflects the
status of the server's hardware components. This health status value can be displayed in several
places in the HP SIM user interface, and is used to access more detailed information about the health
of the system. SNMP and WBEM events, as well as embedded health information from iLO 2, can be
directed to the HP SIM central management server, where actions such as e-mail notifications, paging,
script execution, and event assignment can be automated, and event details can be viewed. In this
example, the system administrator had configured HP SIM to send a text message when certain
production servers exceed a CPU threshold. After HP SIM sends the message, the administrator
consults HP SIM to start diagnosing the problem.
Figure 4 shows an HP SIM table view of a collection of ProLiant servers. The HS (Health Status) column
represents health status for an individual system, and the ES (Event Status) column represents the most
critical, uncleared event that the HP SIM central management server has received for that system. The
SW column reflects recommended updates to ProLiant drivers, firmware, and management agents.
Other status columns reflect additional information in areas such as management processor status and
performance. The Events tab gives quick access to an event viewer and displays a filtered view of the
events for this collection of systems, showing additional event details that might not have been
included in the page notification.
12