4.1.2 Responding to Events That Trigger Traps

This subsection describes all the traps supported by this version of the SNMP agent. These descriptions indicate the event that triggers the trap and include brief suggestions for responding to these events.

It is possible to use the SNMP agent software to monitor boards and to determine if a board has failed. If a board has failed, you can diagnose the board to determine if the hardware is good. If it is, the board can then be restarted. If the diagnosis indicates that the board is bad, then the board must be replaced. If you are using an SNMP manager that can be extended through programming, you can program the manager to diagnose the board when a fault is detected, and either restart the board or notify a person through email or other communication methods.

If there is a hardware fault on your board, the steps described under dlgHiBoardStatusChanged (below) can help you diagnose the issue. This process will not detect a failure in software, but will allow you to restart your system if the software has caused the fault and the hardware is good. Also, if there is a Signal Processor fault, it is possible the POST will not detect this. (Refer to the system release Diagnostics Guide for more information about POST.)

dlgHiBoardStatusChanged

MIB: Hardware Information MIB Module (dlghwinf.mib)

Condition: Change of a board state from OK to failed or from failed to OK

Remarks: This OID identifies the Intel Dialogic or Intel NetStructure board to which it applies by the board's device name. A trap is also sent if the administrative status changes other than by a set. If the board's state changes from OK to failed, perform the following steps:

1. You (or an automated program) should stop the board that has failed. The board is stopped through the use of the dlgHiIdentAdminStatus. This can be found in the dlghwinf.mib. You would set this value to stop pending state (7). This will stop the board.
2. Once the board is stopped, you can run the diagnose function. This is done by setting the dlgHiIdentAdminStatus to diagnose (5). While the board is in the diagnose state, it will not accept any other commands. The diagnose state runs the full power on self test (POST). For more information about POST, refer to the Diagnostics Guide for the system release.
3. Once the POST is complete, a trap will be sent to the SNMP manager. The dlgHiBoardStatus changed trap will be received. If there is an error during POST, it will be sent back in the trap.
4. If the POST did not return an error, then you can start the board. This is done by setting the dlgHiIdentAdminStatus to start-pending. Another trap will be sent up once the board has moved into the Started state.
5. If the POST did return an error, this means the board failed the POST and should be replaced. It should not be started.

   Note: If this process is automated, certain safeguards should be put in place to make sure a board is not restarted too often. For example, if there is a software failure that causes the board to fail immediately after starting, it is possible that the process will continually restart the board. This will cause a large performance hit on the system. When automating the process, it is important to keep a counter on how often a board is restarted. This counter should be checked, and if a board is started too often within a specified amount of time, the restart for that board should be disabled. This will have to be implemented by the programmer as part of the process automation.

dlgHiTestTrap

MIB: Hardware Information MIB Module (dlghwinf.mib)

Condition: Setting of dlgHiIdentTestTrapEnable 0 to a value of 1.

Remarks: This trap can be sent by setting the dlgHiIdentTestTrapEnable OID to a value of 1 (using any MIB browser) to verify that the hardware agent is capable of communicating with the network management station.

dlgDsx1Alarm

MIB: Digital Service Level 1 (DS-1) Line Interface (dlgds1.mib)

Condition: Occurrence of new alarm for T-1 or E-1 interface (alarms include Red (101), Blue (102), Yellow (103), E-1 (104), and end of alarm (100)).

Remarks: In most cases, an alarm indicates a problem with either the network trunk or with the configuration of the Intel Dialogic board's front-end. For example, an alarm might indicate that the Intel Dialogic board is configured to provide clocking, which would conflict with the clocking provided by the network trunk.

dlgIsdnDChanged

MIB: ISDN Configuration and Statistics (dlgisdn.mib)

Condition: Change in the operational status of a D-channel's link access protocol (LAPD).

Remarks: This trap may or may not indicate a problem depending on its value for a given event.

dlgIsdnBChanged

MIB: ISDN Configuration and Statistics (dlgisdn.mib)

Condition: Change in the operational status of a B-channel.

Remarks: This trap may or may not indicate a problem depending on its value for a given event.

Click here to contact Telecom Support Resources