Monitoring your system, although an important part of high-availability,
consumes system resources. You must carefully consider your performance
needs against your need to know as soon as possible when a failure
threatens availability.
System Performance Issues |
 |
The primary performance impact will be related to the polling
interval and the number of resources being monitored. You need to
balance your need to quickly detect failures with your need for
system performance and adjust the number of resources you monitor
and the polling intervals accordingly.
For example, pv_pvlink/status resource comprise the pv_summary
resource. You may only want to create one pv_summary monitoring
request rather than monitoring both pv_summary and pv_pvlink/status
for all disks.
For polling intervals, you may want to set a short interval,
30 seconds, for resources for require quick response after failure,
and set a longer polling interval, 5 minutes or more, for all other
resources.
Network Performance Issues |
|
Although monitoring is not likely to affect network performance,
you may want to make sure that only necessary messages are being
sent. Make sure your monitor requests are configured such that you
are being notified only for things you really need to know.
Printable version
