Polled alerts
A polled alert is generated when the scheduled evaluation of a monitored resource detects an error or other condition in the resource. Polled alerts generally require threshold values to be defined so that OpenEdge Management can track a resource's performance in response to these parameters. For example, threshold values can include defining criteria such as a performance level that is lower or higher than a given number, or identifying the age of a file being older than a particular time (that is, minutes, hours, days, and so forth). Threshold values give you the flexibility to refine rule conditions based on the particular performance values you choose for a resource.
Factors associated with assessing threshold values
OpenEdge Management can trigger an alert when a monitored resource performs outside the currently defined value for a threshold rule. However, the defined polling interval could be set such that OpenEdge Management must complete x number of polling cycles (as defined in the Throw alert after polled field in the Rule definition section of a monitoring plan) before the alert is triggered.
For example, a CPU resource can be polled hourly to see if the CPU usage exceeds 90%. If the percentage does exceed 90%, an alert is generated. However, the alert will not trigger until the number of polling cycles as defined in the Throw alert after polled field is also exceeded.
Alerts that gather data through this polling process allow you to assess data polled over time to determine if your thresholds are too high or too low. You can adjust these threshold values, and the frequency with which the polling cycles are set to occur, at any time to refine your data collection.
The interpretation of a threshold value can vary from resource type to resource type. Differences in these interpretations can also affect what an alert display means for these various resource types. For an explanation of these resource-specific calculations, see the appropriate resource-specific rule details in the system, network, or file resource rule calculations in the OpenEdge Management: Resource Monitoring Guide. For details about database rule details, see the OpenEdge Management: Database Management Guide. For details about OpenEdge Management rules related to the OpenEdge server products, see the OpenEdge Management: Servers Guide.
Example: Reviewing polled alert fields
Figure 1–1 shows the Create Monitoring Plan page for a disk resource. In resource create mode, the Rule definition section contains the Alert if disk activity exceeds field, which is an example of a rule for which a polled alert is generated. The value defined for this threshold, and the values defined for the associated action and alert fields in this same section, enable OpenEdge Management to trigger a polled alert.
![]()
Figure 1–1: Sample polled alert-related data
The Monitoring plan definition section contains the Polling Interval field. Based on the value you define for this field, OpenEdge Management checks (polls) the resource according to the time interval specified. The default value of 900 seconds as shown in Figure 1–1 tells OpenEdge Management to check this disk resource every 15 minutes and keep a count of the number of passed or failed polls. OpenEdge Management also uses this count to assess the alert properties values defined in the Rule definition to determine when to throw or clear an alert. Note that the Alerts Enabled field is selected in Figure 1–1. OpenEdge Management will trigger alerts when the resource is active and a rule is violated.
The Rule definition section shows the rule-, alert-, and action-related fields for a disk resource monitor. Note that the rule definition is actually a combination of the threshold value and the specific action- and alert-related values that identify the activities OpenEdge Management performs if this threshold is exceeded.
As shown in the Rules definition section in Figure 1–1, polled alerts require you to accept default values or set values in a combination of fields to establish the following criteria:
- The specific rule's threshold which, when violated, causes an alert to trigger. In Figure 1–1, the rule threshold is set at 90.0% in the Alert if disk activity exceeds field.
- The number of polls in which the condition must occur before an alert is generated. In Figure 1–1, the Throw alert after field indicates that the alert will be generated after one failed poll.
- Whether to always throw a new alert. In Figure 1–1, the option is selected.
- The number of polls in which the condition must occur without OpenEdge Management detecting any rule violations before the alert is automatically cleared. In Figure 1–1, the value of 0 in the Clear alert after field indicates that the alert will not be automatically cleared. (If you choose not to clear an alert automatically, you must clear it manually.) See the "Clearing alerts through the management console" section for more details about manually clearing alerts.
- The severity of the alert as defined in the Alert severity field: Informational, Warning, Error, or Severe.
- The action, if one is specified, to perform when the alert is generated.
- The action, if one is specified, to perform when the alert is cleared.
For more detailed information about defining alerts, see the "Configuring rules for individual resources" section.
Copyright © 2008 Progress Software Corporation www.progress.com Voice: (781) 280-4000 Fax: (781) 280-4095 |