Juniper Networks, Inc.

Thomas Michael Skibo

Dennis C Ferguson

Chi-Chung Kenny Chen

Devereaux C Chen

Ramesh Padmanabhan

Eric M Verwillow

Pradeep S Sindhu

Jayabharat Boddu

Debashis Basu

Sreeram Veeragandham

Chang-Hong Wu

A network device includes multiple packet processing engines implemented in parallel with one another. A spraying component distributes incoming packets to the packet processing engines using a spraying technique that load balances the packet processing engines. In particular, the spraying component distributes the incoming packets based on queue lengths associated with the packet processing engines and based on a random component. In one implementation, the random component is a random selection from all the candidate processing engines. In another implementation, the random component is a weighted random selection in which the weights are inversely proportional to the queue lengths.

Packet spraying for load balancing across multiple packet processors

A system selectively drops data from a queue. The system includes queues that temporarily store data, a dequeue engine that dequeues data from the queues, and a drop engine that operates independently from the dequeue engine. The drop engine selects one of the queues to examine, determines whether to drop data from a head of the examined queue, and marks the data based on a result of the determination.

Systems and methods for congestion control using random early drop at head of buffer

A transmitting system inserts runt abort packets in an outgoing data stream during idle time inter-frame time fill. The runt abort packets cause the receiving system to synchronize itself to the transmitting system so that even if an error during inter-frame time fill causes the receiving system to go into an erroneous state, the receiving system will be synchronized with the transmitting system before receiving valid data. In one embodiment, the transmitting system transmits data in packets over SONET. The packet data is scrambled at the transmitting end and descrambled at the receiving end. Runt abort packets sent during inter-frame time fill resynchronize the descrambler. If there is an error in the inter-frame time fill bytes, causing the receiving end descrambler to no longer be synchronized with the transmitting end scrambler, the runt abort packets will cause the descrambler to resynchronize state with the transmitting scrambler.

System, apparatus, and method for increasing resiliency in communications

A system updates a cyclic redundancy check (CRC) value. The system receives data containing an arbitrary number of valid and invalid portions. The valid portions are positioned adjacent to one another. The system also receives a signal representing a quantity of valid portions in the data and a current CRC value. The system updates the current CRC value using the data and signal in a single clock cycle.

Single cycle cyclic redundancy checker/generator

Growing community of inventors

Palo Alto, CA, United States of America

Thomas Michael Skibo