Understanding Alert System |
|
An alert is a notable occurrence (possibly automated) that might be of interest to a Process Platform administrator or an operator. Alerts can be detected and described programmatically.
Examples of alerts that are used in Process Platform are:
- Service has started, stopped, or crashed.
- There is an unauthorized access to the system.
- Connection with the database could not be established.
- In general, service containers can go into a problem state. Entering and leaving the problem state is indicated through alerts (error and informational respectively). See Working With Problem Registry.
The Alert System in Process Platform is designed to monitor the health of a software. The Alert System informs an operator or an administrator of a system of undesirable situations such as exceptions, errors, overload on a part of the system, and so on. With such a mechanism in place, the operators can pro-actively respond to deviations in performance and optimize the system usage, thereby preventing the error condition to spread to the other parts of the system. The Alert System provides the functionality to define an alert on a managed component and issue it when the need arises.
The alert system issues, captures, processes, and forwards alerts to various consumers. A consumer is a program that processes the alerts appropriately.
The alert system in Process Platform has the following features:
- Defines the standard structure of alerts
- Provides an interface using which the Process Platform and Process Platform-based applications issue alerts
- Possesses criteria to filter alerts
- Routes the alerts to consumers for specific purposes
- Issues alerts such as JMX notifications
- Manages the descriptive text within the messages
- Provides alerts in default culture-neutral language
An alert can be of the level: ERROR, WARNING, or INFORMATION, depending on the severity of the alert. All the alerts use Log4J category com.eibus.management.AlertSystem.
The Alert Repository for Process Platform describes all the defined alerts and the following information:
| Field | Description | Example |
|---|---|---|
| Problem state related | Indicates whether this alert is related to a problem state. Possible values are:
|
No |
| ID | Unique identification of the alert that is currently only exposed through JMX. | Cordys.ESBServer.Messages.MONITOR_START_ERROR_EXCEPTION_OCCURRED |
| Level | Error, warning, or informational. | Error |
| Message text | English alert message. | The OpenText Process Suite Platform (<instance name>) could not start owing to a catastrophic exception that occurred while starting the same. |
| Description | Description of the alert that is exposed through JMX. | This alert is issued when the OpenText Process Suite Platform (<instance name>) is unable to start because of an exception. |
| Instructions | Instructions for the administrator or operator to correct the situation. | Analyze the Cordys log files and the operating system log to find the exception causing the monitor to fail and resolve that. |
This section covers the following topics: