Abstract: Presented herein are methodologies in which packets or events are selected statistically to update a counter of a network device. The updated value that is stored in the counter also reflects a number of packets (or corresponding bytes) that were not selected to update the counter. The methodology includes receiving, at a network device, a first packet followed by a second packet, probabilistically selecting the second packet to update a value of a counter of the network device while probabilistically not selecting the first packet to update the value of the counter, and updating the value of the counter to account for both the first packet and the second packet.

Abstract: Symbol multiplexing Physical Medium Attachment (PMA) may be provided. A plurality of first lanes may be received and then Alignment Markers (AMs) from the plurality of first lanes may be used to determine symbol boundaries and identify the plurality of first lanes. Next, groups of the plurality of first lanes may be de-skewed and checkerboard patterns in the plurality of first lanes may be undone. Then the plurality of first lanes may be symbol-wise multiplexed to a plurality of second lanes. The plurality of second lanes may then be sent.

Abstract: In one embodiment, a load balancing method may comprise: assigning a plurality of packets of a flow to a plurality of segments according to a segmentation criterion, each segment including one or more packets of said plurality of packets, and at least one of the plurality of segments including more than one packet of the plurality of packets; tagging each packet of the plurality of packets with a segment sequence identifier to indicate to which segment the packet is assigned; and arranging the plurality of packets for transmission via an interconnect so that all packets belonging to a same segment will be transmitted via a same path.

Abstract: A network element is configured to efficiently load balance packets through a computer network. The network element receives a packet associated with flow attributes and generates a Load Balancing Flow Vector (LBFV) from the flow attributes. The network element partitions the LBFV into a plurality of LBFV blocks and reorders the LBFV blocks to generate a reordered LBFV. The LBFV blocks are reordered based on a reordering sequence that is different from reordering sequences on other network elements in the computer network. The network element hashes the reordered LBFV to generate a hash key for the packet and selects a next hop link based on the hash key. The next hop link connects the network elements to a next hop network element in the computer network.

Abstract: A network element is configured to efficiently load balance packets through a computer network. The network element receives a packet associated with flow attributes and generates a Load Balancing Flow Vector (LBFV) from the flow attributes. The network element partitions the LBFV into a plurality of LBFV blocks and reorders the LBFV blocks to generate a reordered LBFV. The LBFV blocks are reordered based on a reordering sequence that is different from reordering sequences on other network elements in the computer network. The network element hashes the reordered LBFV to generate a hash key for the packet and selects a next hop link based on the hash key. The next hop link connects the network elements to a next hop network element in the computer network.

Abstract: In one embodiment, a load balancing method may comprise: assigning a plurality of packets of a flow to a plurality of segments according to a segmentation criterion, each segment including one or more packets of said plurality of packets, and at least one of the plurality of segments including more than one packet of the plurality of packets; tagging each packet of the plurality of packets with a segment sequence identifier to indicate to which segment the packet is assigned; and arranging the plurality of packets for transmission via an interconnect so that all packets belonging to a same segment will be transmitted via a same path.

Abstract: A port extender provides individual flow control for ports multiplexed from a network device. The port extender is configured to multiplex a network device port across the front panel ports on the port extender. The port extender also determines that an input buffer for the network device port is overloaded, and generates a Priority-based Flow Control (PFC) frame identifying that the network device port is congested. The port extender sends the PFC frame via the network device port to the network device. The PFC frame causes the network device coupled to the network device port to discontinue sending data for transmission from the plurality of ports on the port extender.

Abstract: A network element is configured to efficiently load balance packets through a computer network. The network element receives a packet associated with flow attributes and generates a Load Balancing Flow Vector (LBFV) from the flow attributes. The network element partitions the LBFV into a plurality of LBFV blocks and reorders the LBFV blocks to generate a reordered LBFV. The LBFV blocks are reordered based on a reordering sequence that is different from reordering sequences on other network elements in the computer network. The network element hashes the reordered LBFV to generate a hash key for the packet and selects a next hop link based on the hash key. The next hop link connects the network elements to a next hop network element in the computer network.

Abstract: A port extender provides individual flow control for ports multiplexed from a network device. The port extender is configured to multiplex a network device port across the front panel ports on the port extender. The port extender also determines that passing traffic through one of the front panel ports is degraded due to network congestion, and generates a flow control message identifying the front panel port. The port extender sends the flow control message via the network device port to the network device. The flow control message causes the network device coupled to the network device port to discontinue sending data for transmission from the degraded port.

Abstract: In one embodiment, a load balancing method may comprise: assigning a plurality of packets of a flow to a plurality of segments according to a segmentation criterion, each segment including one or more packets of said plurality of packets, and at least one of the plurality of segments including more than one packet of the plurality of packets; tagging each packet of the plurality of packets with a segment sequence identifier to indicate to which segment the packet is assigned; and arranging the plurality of packets for transmission via an interconnect so that all packets belonging to a same segment will be transmitted via a same path.

Abstract: Techniques for flow rate based load balancing are described. In one embodiment, a method includes receiving a packet associated with a packet flow at an ingress top-of-rack switch in a spine-leaf network. The method includes determining identification information for the packet flow, including a source and a destination. Based on the identification information, the method includes determining an instantaneous flow rate for a first path through the network from the source to the destination. The method also includes comparing the instantaneous flow rate to an average flow rate. Based on the comparison between the instantaneous flow rate to the average flow rate, the method includes assigning the packet flow to one of the first path or a second path.

Abstract: Packets that are to be transmitted via a plurality of egress interfaces of a network device are stored in a memory of the network device. The packets are stored in a plurality of queues that respectively correspond to the egress interfaces. The network device determines a set of queues, from among the plurality of queues, for which packet dropping is enabled. The network device determines whether a utilization level of the memory meets a threshold. In response to determining that the utilization level of the memory meets the threshold: the network device randomly or pseudorandomly selects a first queue from the set of queues for which packet dropping is enabled, dequeues a first packet from the selected first queue, and deletes the first packet that was dequeued from the selected first queue.

Abstract: In one embodiment, a load balancing method may comprise: assigning a plurality of packets of a flow to a plurality of segments according to a segmentation criterion, each segment including one or more packets of said plurality of packets, and at least one of the plurality of segments including more than one packet of the plurality of packets; tagging each packet of the plurality of packets with a segment sequence identifier to indicate to which segment the packet is assigned; and arranging the plurality of packets for transmission via an interconnect so that all packets belonging to a same segment will be transmitted via a same path.

Abstract: Techniques for flow rate based load balancing are described. In one embodiment, a method includes receiving a packet associated with a packet flow at an ingress top-of-rack switch in a spine-leaf network. The method includes determining identification information for the packet flow, including a source and a destination. Based on the identification information, the method includes determining an instantaneous flow rate for a first path through the network from the source to the destination. The method also includes comparing the instantaneous flow rate to an average flow rate. Based on the comparison between the instantaneous flow rate to the average flow rate, the method includes assigning the packet flow to one of the first path or a second path.

Abstract: In one embodiment, a load balancing method may comprise: assigning a plurality of packets of a flow to a plurality of segments according to a segmentation criterion, each segment including one or more packets of said plurality of packets, and at least one of the plurality of segments including more than one packet of the plurality of packets; tagging each packet of the plurality of packets with a segment sequence identifier to indicate to which segment the packet is assigned; and arranging the plurality of packets for transmission via an interconnect so that all packets belonging to a same segment will be transmitted via a same path.

Abstract: In one embodiment, a network node comprising: a memory, including one or more memory entries associated with a meter; a sensor adapted to detect a discrepancy between an allocated bandwidth allocated to the meter and a data bandwidth measured by the meter, the allocated bandwidth being a portion of a total allocated bandwidth allocated to a plurality of meters, and the discrepancy being that the allocated bandwidth compared to the data bandwidth is one of: excessive or insufficient; and a generator, wherein the generator is adapted, upon the sensor detecting that the allocated bandwidth is excessive, to generate a message indicative of at least part of the allocated bandwidth being released from the meter, and wherein the generator is further adapted, upon the sensor detecting that the allocated bandwidth is insufficient, to generate a message indicative of a request for an allocation of additional bandwidth to the meter.

Abstract: A data unit is received at a first one of a plurality of network devices associated with a first one of the plurality of link aggregate groups. A hashing function is applied to the data unit to generate a first hash value, where the first hash value identifies a communication link in the first one of the plurality of link aggregate groups. The data unit is received at a second one of the plurality of network devices associated with a second one of the plurality of link aggregate groups. The hashing function is applied to the data unit to generate a second hash value that is distinct from the first value, where the second hash value identifies a communication link in a second one of the plurality of link aggregate groups along which the data unit is to be communicated.

Abstract: A network device determines whether a utilization threshold is reached, the utilization threshold associated with memory resources of the network device, the memory resources including a shared memory and a reserved memory. Available memory in the shared memory is available for any egress interfaces in a plurality of egress interfaces, and the reserved memory includes respective sub-pools for exclusive use by respective egress interfaces among at least some of the plurality of egress interfaces. First packets to be transmitted are stored in the shared memory until a utilization threshold is reached, and in response to determining that the utilization threshold is reached, a second packet to be transmitted is stored in the reserved memory.

Abstract: A network device determines whether a utilization threshold is reached, the utilization threshold associated with memory resources of the network device, the memory resources including a shared memory and a reserved memory. Available memory in the shared memory is available for any egress interfaces in a plurality of egress interfaces, and the reserved memory includes respective sub-pools for exclusive use by respective egress interfaces among at least some of the plurality of egress interfaces. First packets to be transmitted are stored in the shared memory until a utilization threshold is reached, and in response to determining that the utilization threshold is reached, a second packet to be transmitted is stored in the reserved memory.

Abstract: In a network device, a plurality of first counters for measuring respective fill levels of respective queues are maintained, the queues using respective dedicated areas of a memory and a shared area of a memory. A single second counter for measuring a fill level of a shared area is maintained. When a packet is enqueued, the respective first counter corresponding to the queue is updated, and if any part of the packet is stored in the shared area, the single second counter is updated. When the packet is dequeued, the respective first counter is updated, and if any part of the packet was stored in the shared area, the single second counter is updated.