Management of traffic buffering in internal and external memories in a passive optical network

摘要

Methods, systems, and computer program products for managing memory of a passive optical network (PON) are provided herein. The method includes the steps of receiving a plurality of data packets from a stream of packets. A plurality of packet descriptors associated with each data packet is stored in a configurable first queue. The first queue includes a plurality of cache slots and is managed in an internal memory. A state of the first queue is identified. In response to the state of the first queue meeting a predetermined threshold, packet descriptors from a tail cache slot of the first queue are transferred to a second queue. The second queue is managed in an external memory.

主权项

1. A method for managing memory of a passive optical network (PON) comprising:
receiving a plurality of data packets; storing a plurality of packet descriptors associated with each data packet in a configurable first queue, wherein the first queue includes a plurality of cache slots and is managed in an internal memory; adjusting a quantity of the plurality of cache slots based on read and write operations to an external memory; identifying a state of the first queue; in response to the state of the first queue meeting a predetermined threshold, transferring packet descriptors from a tail cache slot of the first queue to a second queue, wherein the second queue is managed in the external memory; and storing additional packet descriptors in the second queue while the state of the first queue meets the predetermined threshold; dequeueing a packet descriptor of the plurality of packet descriptors from the first queue; in response to the dequeuing of the packet descriptor, transferring content of a cache slot from among the plurality of cache slots from the second queue to the first queue, the content comprising at least one of the additional packet descriptors; and changing the state of the first queue when all additional packet descriptors of the second queue are transferred to the first queue and all cache slots of the first queue are available for storage.