Define the following variables for each new alert specification:
Name - A short nickname of the alert specification.
Description - A more detailed description of the alert's purpose and how to use it.
File & Variable - The file/variable that you are monitoring. For a list of each file and its description, see List of Alert Specifications. Check the Override box to specify a custom alert specification. See tutorial below.
Condition & Threshold - The condition and threshold that the monitored variable must meet in order for an alert to be triggered. For example, if you are monitoring "cpu-0 idle" and you set the condition and threshold to "< 20," an alert will be triggered if the cpu-0 idle percentage falls below 20%.
Duration - The duration (in minutes) for which the condition must be true without interruption in order for an alert to be triggered. RightScale evaluates each alert specification once a minute. For example, it may be wise to require the cpu idle to be below 20% for 5 minutes in a row before actually triggering the alert.
Escalation - The name of the alert escalation that will be executed if the alert conditions are met. The escalation list must be defined on the deployment of a server or on the account. A "default" list is used if none matches.
Example: If "cpu-0/cpu-idle" falls below 20% available (< 20) for 5 minutes, then execute the "critical" alert escalation.
New Alert Specification
Alert specifications and escalations are used to configure your deployment to automatically take action on your behalf if certain user-defined conditions are met that trigger an alert. Alerts are commonly used to create scalable server arrays for autoscaling.
Alert Specification - Defines the condition under which an alert is triggered and an alert escalation is called.
Alert - A notification that a problematic condition occurred in the deployment.
Alert Escalation - Defines the action to be taken when an alert is triggered. Typically, each alert specification will have a related alert escalation. That is, if an "alert specification" is met, then it will trigger an "alert" and the related "alert escalation" will be performed on the deployment.
For example, you might want to add another server if the disk usage of your application servers is above 80% for more than an hour. At the same time, you could also send an email notification to your system administrator when an alert has been triggered.
A number of commonly used alert specifications have already been defined for your convenience. You can also create your own custom alert specification. Apply alert specifications to ServerTemplates and individual Servers under the Alerts tab.
Design > Alerts > Specifications