Replication of multicast data packets in a multi-stage switching system

• US 7,519,054 B2
• Filed: 01/27/2005
• Issued: 04/14/2009
• Est. Priority Date: 01/27/2005
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Varma, Anujan
• Primary Examiner(s): Ryman; Daniel J
• Assistant Examiner(s): CRUTCHFIELD, CHRISTOPHER M
• Application Number:
  US11/044,095
• Publication Number:
  US 20060165111A1
• Time in Patent Office: 1,538 Days
• Field of Search: 370/389, 370/351, 370/352, 370/229, 370/388, 370/390
• US Class Current: 370/388
• CPC Class Codes: H04L 49/1515 Non-blocking multistage, e.g. Clos
• CPC Class Codes: H04L 49/201 Multicast operation; Broadcast operation
• CPC Class Codes: H04L 49/3018 Input queuing
• CPC Class Codes: H04L 49/357 Fibre channel switches
• CPC Class Codes: H04Q 3/68 Grouping or interlacing selector groups or stages

Abstract

In general, in one aspect, the disclosure describes a multi-stage switch having at least one ingress switch module to receive data and to generate frames that are transmitted as a wavelength division multiplexed signal. The multi-stage switch further includes a core switch module operatively connected to receive the wavelength division multiplexed signal from the at least one ingress switch module and to switch the frames. The multi-stage switch additionally includes at least one egress switch module to receive the wavelength division multiplexed signal from the core switch module and to transmit data. The at least one ingress switching module and the at least one egress switching module are capable of replicating multicast data packets.

Claims

1. A multi-stage switch comprising:

at least one ingress switching module to receive data packets, to switch the data packets, to generate frames from the switched data packets, and to form wavelength division multiplexed signals from the frames, wherein said at least one ingress switch module comprises at least one ingress queuing engine operatively connected to transmit the data packets to at least one crossbar switch plane, which is operatively connected to switch the data packets to at least one ingress crossbar data element, wherein data transfer is controlled by an ingress scheduler, wherein said at least one ingress switch module further includes a multicast lookup table, the multicast lookup table used to identify egress switching modules associated with a multicast data packet, wherein the multicast lookup table is also used to determine type of replication associated with the multicast data packet, wherein multicast data packets are assigned a multicast identification, and wherein the multicast lookup table identifies the associated egress switching modules and the type of replication to be used based on the multicast identification;

a core switching module operatively connected to receive the wavelength division multiplexed signals from the at least one ingress switching module and to switch the wavelength division multiplexed signals; and

at least one egress switching module to receive the wavelength division multiplexed signals from the core switching module, to extract the frames from the wavelength division multiplexed signals, to extract the data packets from the frames, to switch the data packets, and transmit the data packets,

wherein said at least one ingress switching module and said at least one egress switching module are capable of replicating multicast data packets.

2. The switch of claim 1, wherein a particular ingress switching module distributes data packets destined for a particular egress switching module across the at least one crossbar switch plane such that the data packets destined to the particular egress switching module are distributed as evenly as possible across the at least one ingress crossbar data element.

3. The switch of claim 2, wherein data packets queued in different ones of the at least one ingress crossbar data element associated with the particular ingress switching module and destined to the particular egress switching module are switched in parallel across the core switching module to the particular destination egress switching module.

4. The switch of claim 1, wherein said at least one egress switch module includes a set of internal sub-modules and the particular egress switch module uses the set of internal sub-modules to receive the data arriving from the core switch module.

5. The switch of claim 1, wherein an ingress switching module will replicate a multicast data packet if the multicast data packet is to be delivered from the ingress switching module to at least two egress switching modules.

6. The switch of claim 1, wherein the ingress queuing engine replicates the multicast data packets and transmits copies of the multicast data packets in consecutive frame periods if serial replication is identified.

7. The switch of claim 1, wherein the ingress queuing engine transmits the multicast data packet to the at least one crossbar switch plane and the at least one crossbar switch plane transmits the multicast data packet to multiple ingress crossbar data elements associated with the identified egress switching modules during a frame period if spatial replication is identified.

8. The switch of claim 1, wherein said at least one ingress switch module further includes a framer and wavelength division multiplexer transmitter to form frames from stored segments and create a wavelength division multiplexed optical signal of the frames from the at least one ingress crossbar data elements.

9. The switch of claim 1, wherein the ingress scheduler controls data flow by arbitrating amongst received requests.

10. The switch of claim 9, wherein the requests associated with multicast data packets include a new multicast request field that is active until a first grant for the multicast data packet is received and is used for error detection and recovery of grants.

11. The switch of claim 1, wherein an egress switching module will replicate a multicast data packet if the multicast data packet is to be delivered from the egress switching module to at least two destinations.

12. The switch of claim 1, wherein said at least one egress switch module comprises at least one egress crossbar data element, operatively connected to transmit the data packets to at least one crossbar switch plane, which is operatively connected to switch the data packets to at least one egress queuing engine, and an egress scheduler to control data transfer therethrough.

13. The switch of claim 12, wherein said at least one egress switch module further includes a multicast data lookup table, the multicast lookup table used to determine at least some subset of destinations associated with said at least one egress switch module and type of replication to be used for the multicast data packet.

14. The switch of claim 12, wherein said at least one egress switch module further includes a wave division receiver and deframer to receive a wavelength division multiplexed signal from said core switching module and extract frames therefrom.

15. A method comprising:

receiving data packets at a multistage switch, wherein the multistage switch includes a plurality of ingress switching modules, a core switching module operationally connected to the plurality of ingress switching modules, and a plurality of egress switching modules operationally connected to the core switch module, and wherein the data packets received by the ingress switch modules may include multicast packets that are destined for at least two destinations;

for multicast packets, utilizing a multicast lookup table within the ingress switching modules to identify which egress switching modules a multicast data packet is destined for and what type of replication should be used based on a multicast identification assigned to the multicast data packet, replicating the multicast packet for the at least two destinations, wherein said replicating can be distributed between receiving ingress switching module and egress switching modules associated with the at least two destinations;

within the ingress switching modules, switching the data packets, aggregating the data packets destined for different egress switching module into frames, generating wavelength-division multiplexed (WDM) optical signals from the frames, and transmitting the WDM optical signals to the core switching module;

switching the WDM optical signals through the core switch module via multiple channels;

within the egress switching modules, receiving the WDM optical signals, extracting frames from the WDM optical signals, extracting the data packets from the frames, switching the data packets, and transmitting the data packets to destinations.

16. The method of claim 15, wherein said replicating includes replicating the multicast packet in the ingress switching module if the multicast packet is associated with more than one egress switching module.

17. The method of claim 16, wherein number of times the multicast data packet is replicated is based on number of associated egress switching modules.

18. The method of claim 15, wherein said replicating includes replicating the multicast packet in an associated egress switching module if the multicast packet is associated with more than one destination for the associated egress switching module.

19. The method of claim 18, wherein number of times the multicast data packet is replicated is based on number of destinations associated with the associated egress module.

20. The method of claim 15, wherein said replicating includes spatial replicating where the replication and transmission of the multicast data packet is performed during a single frame period by utilizing multiple crossbar planes.

21. The method of claim 15, wherein said replicating includes serial replication where the replication and transmission are performed one copy at a time during consecutive frame periods.

22. The method of claim 15, wherein multicast packets received are assigned a multicast identification, and further comprising utilizing a multicast lookup table to identify parameters associated with the multicast packet based on the assigned multicast identification.

23. The method of claim 22, wherein a parameter associated with the multicast identification is identity of egress switching modules associated with the multicast data packet.

24. The method of claim 22, wherein a parameter associated with the multicast identification is type of replication to be performed for the multicast data packet.

25. The method of claim 22, wherein a parameter associated with the multicast identification is identity of destinations associated with the multicast data packet.

26. The method of claim 15, further comprising utilizing a multicast lookup table within the egress switching modules to identify which destinations a multicast data packet is destined for and what type of replication should be used based on a multicast identification assigned to the multicast data packet.

27. The method of claim 15, further comprising

activating a new multicast request field within a request frame of a multicast data packet until a first grant is received for the multicast data packet, and

utilizing the new multicast field to detect and correct errors in multicast grants.