PBX Systems Management and Administration: System Diagnostics and Maintenance

The primary objective of system maintenance is to detect, report, and clear trouble as quickly as possible with minimum disruption of service. Periodic tests, automatic software diagnostic programs, and fault detection hardware allow most troubles to be traced to an individual assembly in the system. System diagnostic functions include:

• Monitoring of processor status

• Monitoring and testing of all port and service circuit packs

• Monitoring and control of power units, fans, and environmental sensors

• Monitoring of peripherals (voice terminals and trunk circuits)

• Initiating emergency transfer and control to backup systems

• Originatng alarm information and activate alarms

There is a specific maintenance strategy and plan for each of these hardware elements monitored by the system.

The maintenance subsystem software is responsible for initializing and maintaining the system. This software continuously monitors the system and maintains a record of detected errors. The maintenance subsystem also provides a user interface for on-demand testing and contains two general categories: system alarm troubles that are automatically reported to a local maintenance terminal or a remote maintenance center and user-reported troubles resulting from service problems at individual station user terminals.

The major part of maintenance is system-alarmed troubles. PBX system diagnostic circuitry detects and reports most problems automatically. When the trouble is repaired and no longer detected, the alarm is automatically retired. It is not necessary for personnel to retire alarms after a problem is corrected. Dedicated maintenance circuit packs or daughterboards are used in fault detection and repair at many system levels, including the common control complex, expansion cabinets and carriers, and a variety of trunk interface cards, particularly those used for digital trunk connections. Almost all circuit packs have LED indicators to indicate alarm conditions (red) if the system has detected a fault in that circuit pack. A yellow alarm condition indicates the system is running tests on that circuit pack, and a green condition indicates that the circuit pack is operating without problem. In-line error detection circuitry checks for correct operation.

Maintenance tests can be periodic or on demand. Periodic tests run automatically at fixed intervals on a specific schedule. Usually, short tests are run hourly or less; long tests are run every 24 hours. Demand tests are run by the system when it detects a need or by personnel when required during trouble-clearing activities. Demand tests include the periodic tests, and other tests are required only when trouble occurs. Some nonperiodic tests may be disruptive to system operation. From a terminal, personnel can initiate the same tests the system initiates, and the results are displayed on the terminal screen.

If any part of the system fails any portion of the periodic tests a preset number of times, the system automatically generates an alarm. There are three alarm types common to most systems:

1. Major alarms—Failures that cause critical degradation of service and require immediate attention.

2. Minor alarms—Failures that cause marginal degradation of service but do not render a crucial portion of the system inoperable. This condition requires action, but its consequences are not immediate. Problems that cause minor alarms might be impaired service in a few trunks or stations or interference with one feature across the entire system.

3. Warning alarms—Failures that are localized and cause no noticeable degradation of service. Warning alarms are not reported to the attendant console or a remote location.

The PBX system can usually send an alarm to any customer device such as a light, automatic dialer, a bell, or other equipment. The alarm activation level field on the system parameters maintenance screen must be administered to indicate the alarm level (major, minor, warning, or none) that activates the alarm device.

If the maintenance software detects an error condition related to a specific maintenance element, the system will automatically attempt to repair a problem or operate around it. If a hardware component incurs too many errors, an alarm is generated. Records of each error and alarm are stored. The error log is a record of system errors and can be accessed from a SAT. The error log is useful for analyzing problems that have not caused an alarm or when alarms cannot be retired by replacement of hardware. When errors result in alarms, the alarms are listed in the alarm log. This log can be displayed on a terminal. If several alarms are active, the alarm log can be used to determine the alarms that should be cleared first.