Method for Managing Forwarding Table of Network Processor according to Address Learning in the Linecard Processor composing of Multiple Network Processors and Multiplexing/Demultiplexing Module

摘要

the present invention the plurality of network processors and a multiplexer / demultiplexer of the line card in an Ethernet system having a plurality single by adjusting the repetitive learning results by reducing the forwarding address table access frequency of each network processor with one, relates to the forwarding table entry update and delete overhead of each network processor in a forwarding table of the network management method of reducing a time processor . ; The present invention, communication with the main processor, and each having a plurality of network processors and a plurality of port multiplexer / demultiplexer in the forwarding table of the network management method according to the address learning in the line card processor having In case the Ethernet data frames with the same source address to the line card is received repeatedly up to N times, demultiplexer sequentially line card frame transmission step for transmitting the frame to each port of each network processor; Each network processor is receiving the frame and adding an entry to check their address learning table, you reset the aging timer (aging timer), the main processor frame transmission step for transmitting said received frame to the main processor; A main processor having received the frame is a duplicate entry confirmation step of confirming that the same entry exist in their forwarding tables by using the source address; And if there is the confirmation that the same entry, LRNG_CNT check the flag value to increase by 1 the value is greater than zero LRNG_CNT flag, when the same entry is not present, and to add the entry in the forwarding table, LRNG_CNT flag value a first increased and then, a request to create an entry step for an entry in the forwarding table created in the respective network via a line card processor. ; According to the present invention, a plurality of network processors as the number of interfaces can be reduced in that the main processor and forwarding table update, or deletion occurs in the respective network processor to 1.