Designing and Managing Behavior Models - Communications and Data -
Previous: More about Exporting Objects     Next: Appendix: Debugging a Behavior Model

Appendix: Communications and Data

As a tool that comprehensively monitors and manages your network, NerveCenter uses a variety of data transfers to gather, correlate, disseminate, and store information about network events. This appendix outlines the general flow of data into, through, and out of NerveCenter in the course of its operation.

NerveCenter's primary sources of network information are SNMP traps and device responses to NerveCenter polls. If configured appropriately, Open NerveCenter responds to trap and poll data by forwarding it to your network management platform and to other NerveCenters. For example, forwarded event data might ultimately land in a network management platform's Event Categories window or trigger an alarm transition in a central NerveCenter. Although this sequence may happen quickly, the actual communication path from initial receipt of trap or poll data to the final event message has many stages.

As the first illustration shows, a trace of the communication path initiated by a managed device's SNMP trap or poll response might look like this:

1. Traps are relayed directly to the Open NerveCenter Server if the platform and the server are running on different machines. If they're running on the same machine, traps are detected by the operating system trap service or the management platform's trap multiplexor, such as trapd, and then forwarded to the Open NerveCenter trapper process. The trapper process, in turn, forwards the trap to Open NerveCenter.
2. Open NerveCenter trap masks filter incoming traps to see if they are of interest. If a trap is of interest, an internal event, called a trigger, is generated and used by active alarms. Polls evaluate the poll data returned by managed devices and also use triggers to pass data to alarms.
3. Open NerveCenter alarms correlate the traps and polls with other related data. For example, an alarm might detect that this is the third trap of the same type from the same machine. The alarm then takes any automated actions that were associated with this trap detection. For example, it could issue a trouble ticket or change the device configuration.

   Data Flow

   

   Click the thumbnail above to view full-sized image.

4. If an alarm transition contains the Inform action, the alarm sends a message to the Open NerveCenter platform adapter process, which always resides on the same host as the network management platform, and/or to any listed NerveCenters.
5. The platform adapter determines whether the message requires changing a symbol's color on the map, initiating an event message, or both. Messages to other NerveCenters forward the trap data.
6. If color changes are required, the platform adapter sends a message to the Open NerveCenter ncapp process, which in turn forwards instructions for color changes to the platform map with an API.
7. If an event is to be posted, the platform adapter uses an API to submit a data structure that resembles an SNMP trap to the platform event facility, which decodes traps, associates text messages with events, and posts them in the Event Categories window.

NerveCenter is a client/server application. The NerveCenter server acts as the hub for the data transfers described in this appendix. As shown in the following illustration, event information moves from managed device to NerveCenter server to management platform. But data also flows between the server and other NerveCenter components in support of this flow.

NerveCenter Components

The components shown in the preceding figure are defined in the following table:

NerveCenter Components

The following figure shows the utilities that install NerveCenter and assist in database management:

Utilites for Installation and Database Management

The utilities shown in the preceding figure are defined in the following table.

NerveCenter Utilities