US 12,273,270 B2
Congestion management techniques
Jeongkeun Lee, Mountain View, CA (US); Rong Pan, Saratoga, CA (US); Changhoon Kim, Palo Alto, CA (US); Jeremias Blendin, Santa Clara, CA (US); Georgios Nikolaidis, Mountain View, CA (US); Ashutosh Agrawal, Fremont, CA (US); Ethan M. Spiegel, Mountain View, CA (US); and Anurag Agrawal, Santa Clara, CA (US)
Assigned to Intel Corporation, Santa Clara, CA (US)
Filed by Intel Corporation, Santa Clara, CA (US)
Filed on May 19, 2020, as Appl. No. 16/878,466.
Claims priority of provisional application 62/967,003, filed on Jan. 28, 2020.
Prior Publication US 2020/0280518 A1, Sep. 3, 2020
Int. Cl. H04L 47/11 (2022.01); H04L 47/12 (2022.01); H04L 47/263 (2022.01); H04L 47/30 (2022.01)
CPC H04L 47/115 (2013.01) [H04L 47/12 (2013.01); H04L 47/263 (2013.01); H04L 47/30 (2013.01)] 24 Claims
OG exemplary drawing
 
1. A method comprising:
a network element receiving a packet and
the network element:
determining a probability that a target queue will be congested prior to enqueuing the packet into the target queue,
generating a congestion notification packet for transmission and transmitting the congestion notification packet to an endpoint sender of the packet before storing the packet into a target queue, wherein the congestion notification packet comprises (1) an estimated pause time associated with reducing a level of the target queue to a queue depth and (2) a destination Internet Protocol (IP) address of the packet, wherein the generating the congestion notification packet is based on the probability that the target queue will be congested, and wherein the storing the packet into the target queue comprises storing the packet into the target queue that is determined to be congested.