Congestion notification across multiple layer-2 domains

1. A computing system, comprising:

a processor;

a computer-readable storage medium storing instructions which when executed by the processor causes the processor to perform a method, the method comprising:

removing a layer-2 header of a congestion notification message that includes a destination layer-2 address set as the computing system's layer-2 address, wherein the congestion notification message includes a sampled packet from a congested flow; and

assigning a layer-2 destination address to a locally generated second layer-2 header for the congestion notification message, wherein the assigned layer-2 destination address is derived based on the source address of the sampled packet.

2. The computing system of claim 1 , wherein a source layer-2 address and optionally a virtual local area network (VLAN) identifier of the congestion notification message are modified based on the source address of the sampled packet from the congested flow.

3. The computing system of claim 1 , wherein the source address of the sampled packet from the congested flow is contained in a media access control (MAC) service data unit in the congestion notification message.

4. The computing system of claim 1 , wherein the source address of the sampled packet from the congested flow is an Internet Protocol (IP) address or Fibre Channel source identifier.

5. The computing system of claim 1 , wherein the congestion notification message includes a time-to-live field.

6. The computing system of claim 1 , wherein the method further comprises:

determining a layer-2 destination address based on the source address of the sampled packet from the congested flow; and

generating the second layer-2 header based on the determined layer-2 destination address for the congestion notification message.

7. The computing system of claim 1 , wherein the layer-2 header is an Ethernet header.

8. The computing system of claim 1 , wherein the method further comprises looking up a destination layer-2 address based on the source address of the sampled packet from the congested flow.

9. The computing system of claim 1 , wherein the received congestion notification message is in compliance with Institute of Electrical and Electronics Engineers (IEEE) standard 802.1Qau; and

wherein the source address of the sampled packet from the congested flow is contained in a MAC service data unit within the message.

10. A computer executable method, comprising:

removing a layer-2 header of a congestion notification message that includes a destination layer-2 address set as a local computing system's layer-2 address, wherein the congestion notification message includes a sampled packet from a congested flow; and

assigning a destination layer-2 address to a locally generated second layer-2 header for the congestion notification message, wherein the assigned layer-2 destination address is derived based on the source address of the sampled packet.

11. The method of claim 10 , further comprising modifying a source layer-2 address and optionally a virtual local area network (VLAN) identifier of the congestion notification message, based on the source address of the sampled packet from the congested flow.

12. The method of claim 10 , wherein the source address of the sampled packet from the congested flow is contained in a media access control (MAC) service data unit in the congestion notification message.

13. The method of claim 10 , wherein the source address of the sampled packet from the congested flow is an Internet Protocol (IP) address or Fibre Channel source identifier.

14. The method of claim 10 , further comprising decrementing a time-to-live field included in the congestion notification message.

15. The method of claim 10 , further comprising:

determining a layer-2 destination address based on the source address of the sampled packet from the congested flow; and

generating the second layer-2 header based on the determined layer-2 destination address for the congestion notification message.

16. The method of claim 10 , wherein the layer-2 header is an Ethernet header.

17. The method of claim 10 , further comprising looking up a destination layer-2 address based on the source address of the sampled packet from the congested flow.

18. The method of claim 10 , wherein the received congestion notification message is in compliance with Institute of Electrical and Electronics Engineers (IEEE) standard 802.1Qau; and

wherein the source address of the sampled packet from the congested flow is contained in a MAC service data unit within the message.

19. A switch, comprising:

a header processing module operable to removing a layer-2 header of a congestion notification message that includes a destination layer-2 address set as the switch's layer-2 address, wherein the congestion notification message includes a sampled packet from a congested flow; and

a header-generating module operable to assign a layer-2 destination address to a locally generated second layer-2 header for the congestion notification message, wherein the assigned layer-2 destination address is derived based on the source address of the sampled packet.

20. The switch of claim 19 , wherein the source address of the sampled packet from the congested flow is deduced based on information included in an encapsulated frame; and

wherein the encapsulated frame is a Fibre Channel over Ethernet (FCoE) frame, and wherein the FCoE frame includes an Ethernet header, an FCoE header, and an encapsulated Fibre Channel (FC) frame.

21. The switch of claim 20 , further comprising an FC media-access-control address prefix (FC-MAP) lookup module operable to look up an FC-MAP based on a virtual local area network (VLAN) ID included in the FCoE frame.

22. The switch of claim 21 , wherein the source address of the sampled packet from the congested flow is deduced by appending an FC source identifier (FC S_ID) included in the encapsulated FC frame to the FC-MAP.

23. The switch of claim 19 , wherein the second layer-2 header includes an Ethernet destination address field, an Ethernet source address field, and a virtual local area network (VLAN) identifier field.

24. The switch of claim 19 , wherein the switch is a routing bridge (RBridge) which implements a Transparent Interconnection of Lots of Links (TRILL) protocol, and wherein the congestion notification message is encapsulated in a TRILL packet.