Abstract: Certain aspects of a method and system for Fibre Channel and Ethernet interworking may include receiving by a central processing unit (CPU), one or more requests for discovery of one or more Fibre Channel (FC) switches based on a Fibre Channel initialization protocol (FIP). The CPU may be operable to facilitate, via a network switch, a proxy login into the one or more FC switches based on the FIP for one or more servers. The one or more servers may be operable to directly communicate and/or receive packets to/from the FC switches via the network switch. The network switch may be operable to determine a destination address of one or more servers based on the facilitated proxy login by the CPU if the one or more servers utilize fabric provided MAC addresses (FPMA) or utilize a MAC address table if the one or more servers utilize server provided MAC addresses (SPMA).

Abstract: An Ethernet link may comprise a plurality of active and/or silent channels. Training of one or more link partners for one or more silent channels may be triggered based on monitoring training parameters for an active channel. The link partners may be trained for the silent channels and link partners corresponding to the silent channels may be configured according to the training. Training parameters for an active channel may be compared to a threshold or relative change value. The threshold or relative change value may be fixed, may be based on measurement results and/or may be programmable. The link partners may be configured for silent channels for an echo canceller, a far-end crosstalk canceller and/or a near-end crosstalk canceller. The silent channels may be trained based on transmitting a pulse or LDPC frame via the silent channels.

Abstract: It may be determined via monitoring whether a first link partner and/or a second link partner coupled via an Ethernet link has trained or refreshed circuitry and/or has updated at least one parameter. Based on the determination, a corresponding link partner may be trained, refreshed and/or updated. One or more of an echo canceller, a far-end crosstalk canceller and a near-end crosstalk canceller for one or more channels may be configured based on the training, refreshing and/or updating. One or more channels may be silent and/or one or more may be active. Link partners may communicate via one or more of in-band signaling, out-of-band signaling to determine which link partner may monitor and/or which may control or initiate operation. The controlling link partner may be assigned a master mode of operation. Training, refreshing and/or updating for the monitoring link partner may be based on a timer.

Abstract: Certain aspects of a method and system for minimum frame size support for a communication protocol encapsulated over Ethernet may include an encapsulated protocol over Ethernet (EPoE) frame comprising a start of padding (SP) field. In the EPoE frame, a size of an encapsulated protocol frame that is less than a size of an Ethernet frame may be adjusted by insertion of one or more padding bits so that the size of the encapsulated protocol frame is equal to the size of the Ethernet frame. The SP field may indicate a starting location of the insertion of one or more padding bits within the EPoE frame. The SP field comprises an offset value that may be relative to a location of a word in which the SP field is located or is relative to a location of a word corresponding to a fixed offset from a start of the encapsulated protocol frame.

Abstract: A data communication device configured to communicate with a link partner is provided. The data communication device comprises a port coupled to link partner and an auto-negotiation system configured to generate a first differential Manchester signal that comprises information that identifies at least one mode of operation of the data communication device and configured to transmit the first differential Manchester signal to the link partner using the port.

Abstract: Certain aspects of a method and system for Ethernet congestion management may include transmitting one or more data packets in a network flow from a source node to a destination node. The source node may be enabled to receive one or more congestion notification frames based on detecting congestion at one or more switches in the network flow. A probe packet comprising a departure timestamp, Ts, may be transmitted from the source node to the destination node based on receiving one or more congestion notification frames. The arrival rate, Ra, of one or more data packets at the destination node may be calculated based on an arrival timestamp, Ta, inserted into the probe packet at the destination node. A rate limiter at the source node may be enabled to limit a rate of one or more data packets based on the calculated arrival rate, Ra.