Raritan COMMANDCENTER NOC, Administrator'S Manual

Page: 33 / 156

C
HAPTER
2:
GENERAL AND ADVANCED ADMINISTRATION
21
Configure Performance Thresholds
This page displays the current values at which SNMP performance metrics are considered
problematic and events are generated. You have complete control over these thresholds, including
their value, their re-arm values, and the number of consecutive data samples, for example,
"triggers" which must be exceeded before an event is generated.
Note: Performance thresholds are configured for devices with Infrastructure, Server, or
Promoted Workstation licenses.
By configuring performance thresholds, administrators can adjust the high/low thresholds of
certain SNMP performance metrics. This function puts considerable configuration power into the
hands of the administrator – but one that should be used only in the event of clear evidence of
need, for example, environments that have servers that run under a heavy load constantly might
want to increase the high threshold – but only after receiving alarms that do not indicate a
problem. Setting thresholds too low or too high can result in either too many notifications or a
lack of notifications for problems respectively. Thresholds can also be configured on a per-device
basis – see
Edit Performance Thresholds (Per Device)
later in this chapter. Per-device
thresholds override values that are configured here.
Please see Appendix C: Performance Monitoring for additional information.
1. Click on the
Admin
tab in the top navigation bar.
2. Click
Network Management Configuration
.
3. Click
Configure Performance Thresholds
.
Figure 15 Configure Performance Thresholds
The
Value
column indicates the threshold, which varies by metric, at which an alarm condition of
either “high” or “low” exists, depending upon the metrics of that performance indicator. The
Rearm At
column is an indicator of the value at which the threshold alarm will reset, after it has
detected an alarm condition. The
Trigger
column details how many polling cycles the value must
be above or below the Value level to trigger an alarm. Example: CPU Utilization set at 95.0
Value, 50.0 Rearm, and Trigger at 3 would issue a “High Threshold” alarm if the node had 95% or
higher CPU usage over 15 minutes (3 5-minute intervals). If the Value dropped below 50% either
during or after the initial Trigger, then the alarm would reset and a new alarm would be issued if
the same conditions reappeared.
Note: If the threshold is of type High, the value must be greater than or equal to the Rearm At
value. If the threshold is of type Low, the value must be less than or equal to the Rearm At value.