Intelligent feed switch

• US 10,872,078 B2
• Filed: 05/31/2018
• Issued: 12/22/2020
• Est. Priority Date: 03/27/2012
• Status: Active Grant

• Current Assignee: Exegy Incorporated
• Original Assignee: Ip Reservoir, LLC
• Inventors: Parsons, Scott, Taylor, David E., Indeck, Ronald S.
• Primary Examiner(s): Oyebisi, Ojo O
• Application Number:
  US15/994,262
• Publication Number:
  US 20180276271A1
• Time in Patent Office: 936 Days
• Field of Search: 705 37
• CPC Class Codes: G06F 16/24 Querying
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
• CPC Class Codes: H04L 45/72 Routing based on the source address

Abstract

Various techniques are disclosed for offloading the processing of data packets. For example, incoming data packets can be processed through an offload processor to generate a new stream of outgoing data packets that organize data from the data packets in a manner different than the incoming data packets. Furthermore, in an exemplary embodiment, the offloaded processing can be resident in an intelligent switch, such as an intelligent switch upstream or downstream from an electronic trading platform.

Claims

1. A packet switching method comprising:

receiving a plurality of incoming feed-specific data packets via at least one port of a switch, the incoming feed-specific data packets corresponding to a plurality of different data feeds, the incoming feed-specific data packets comprising payloads that include a plurality of messages, the messages comprising message data, the switch comprising (1) a plurality of ports, (2) switching logic, and (3) a processor, wherein the switching logic and processor are co-resident within the switch;

filtering and sorting, by the processor, the message data based on data consumer-specific criteria with respect to interests of a plurality of data consumers to select a plurality of the messages according to the interest criteria;

generating, by the processor, a plurality of customized outgoing data consumer-specific data packets by repacketizing the selected messages into the outgoing data consumer-specific data packets such that each outgoing data consumer-specific data packet comprises newly organized payloads of messages that were commonly selected according to the interest criteria;

determining, by the switching logic, at least one of the ports for a plurality of the outgoing data consumer-specific data packets; and

outputting the outgoing data consumer-specific data packets via the determined at least one port.

2. The method of claim 1 wherein the filtering and sorting step comprises the processor performing a plurality of data consumer-specific search operations on the message data with respect to a plurality of search terms, the search terms being associated with the data consumers to find message data of interest to the data consumers.

3. The method of claim 2 wherein the search operations comprise at least one member of the group consisting of an exact matching operation, an approximate match operation, and a regular expression pattern match operation.

4. The method of claim 2 further comprising the processor, based on the search operations, storing the message data found to be of interest to the data consumers in a plurality of queues, each queue being associated with a data consumer such that the message data found to be of interest to a particular data consumer is stored in the queue associated with that data consumer.

5. The method of claim 4 wherein the repacketizing step comprises the processor generating the outgoing data consumer-specific data packets from commonly-queued message data.

6. The method of claim 5 wherein the repacketizing step comprises the processor (1) selecting a queue from which to generate an outgoing data consumer-specific data packet, (2) accessing packaging parameter data that is associated with the selected queue, and (3) generating an outgoing data consumer-specific data packet from message data in the selected queue in accordance with the accessed packaging parameter data.

7. The method of claim 1 further comprising the processor encrypting at least a portion of the messages.

9. The method of claim 1 further comprising the processor normalizing at least a portion of the messages.

10. The method of claim 9 wherein the normalizing step comprises the processor performing different data consumer-specific normalization operations on the message data.

11. The method of claim 1 wherein the data feeds include at least one social network data feed.

12. The method of claim 1 wherein the data feeds include at least one content aggregation feed.

13. The method of claim 1 wherein the data feeds include at least one machine-readable news feed.

14. The method of claim 1 further comprising:

the switch performing a protocol transformation to generate the outgoing data consumer-specific data packets such that the outgoing data packets exhibit a different protocol than the received data packets.

15. The method of claim 1 wherein the processor comprises a field programmable gate array (FPGA), the FPGA comprising pipelined firmware logic, the pipelined firmware logic performing the filtering and sorting and generating steps in parallel with respect to successively received incoming data packets.

16. The method of claim 1 wherein the switching logic is resident on the processor.

17. The method of claim 1 wherein the processor comprises a field programmable gate array (FPGA).

18. The method of claim 17 wherein the switching logic is resident on another FPGA.

19. The method of claim 18 further comprising the FPGAs communicating with each other via a member of the group consisting of (1) a custom interface, (2) a PCI-express interface, (3) a XAUI interface, and (4) an Ethernet interface.

20. The method of claim 1 wherein the switching logic is resident on an application specific integrated circuit (ASIC).

21. The method of claim 20 wherein the processor comprises a field programmable gate array (FPGA), the method further comprising the ASIC and the FPGA communicating with each other via a member of the group consisting of (1) a custom interface, (2) a PCI-express interface, (3) a XAUI interface, and (4) an Ethernet interface.

22. The method of claim 1 wherein the processor comprises at least one member of the group consisting of a reconfigurable logic device, a graphics processor unit (GPU), and a chip multi-processor (CMP).

23. The method of claim 1 wherein at least a plurality of the received data packets comprise transmission control protocol (TCP) data packets, the method further comprising the processor performing a TCP termination on at least a plurality of the received TCP data packets.

24. The method of claim 22 wherein the at least one member comprises a processing pipeline, the method further comprising:

the processing pipeline performing the filtering and sorting and generating steps in a pipelined fashion as the message data streams through the at least one member.

25. The method of claim 24 wherein the filtering and sorting step comprises the processing pipeline performing a plurality of data consumer-specific search operations on the message data with respect to a plurality of search terms, the search terms being associated with the data consumers to find message data of interest to the data consumers.

26. The method of claim 25 further comprising the processing pipeline, based on the search operations, storing the message data found to be of interest to the data consumers in a plurality of queues, each queue being associated with a data consumer such that the message data found to be of interest to a particular data consumer is stored in the queue associated with that data consumer.

27. The method of claim 26 wherein the repacketizing step comprises the processing pipeline generating the outgoing data consumer-specific data packets from commonly-queued message data.

28. The method of claim 27 wherein the repacketizing step comprises the processing pipeline (1) selecting a queue from which to generate an outgoing data consumer-specific data packet, (2) accessing packaging parameter data that is associated with the selected queue, and (3) generating an outgoing data consumer-specific data packet from message data in the selected queue in accordance with the accessed packaging parameter data.

8. The method of 7 wherein the encrypting step comprises the processor performing different data consumer-specific encryption operations on the message data.

29. A packet switch comprising:

a plurality of ports, wherein at least one of the ports is for receiving a plurality of incoming feed-specific data packets, the incoming feed-specific data packets corresponding to a plurality of different data feeds, the incoming feed-specific data packets comprising payloads that include a plurality of messages, the messages comprising message data, switching logic; and

a processor;

wherein the switching logic and processor are co-resident within the packet switch;

wherein the processor is configured to (1) filter and sort the message data based on data consumer-specific criteria with respect to interests of a plurality of data consumers to select a plurality of the messages according to the interest criteria, and (2) generate a plurality of customized outgoing data consumer-specific data packets by repacketizing the selected messages into the outgoing data consumer-specific data packets such that each outgoing data consumer-specific data packet comprises newly organized payloads of messages that were commonly selected according to the interest criteria; and

wherein the switching logic is configured to determine at least one of the ports for outputting a plurality of the outgoing data consumer-specific data packets.

30. The packet switch of claim 29 wherein the processor comprises at least one member of the group consisting of a reconfigurable logic device, a graphics processor unit (GPU), and a chip multi-processor (CMP).

31. The packet switch of claim 29 wherein the processor comprises a field programmable gate array (FPGA).

32. The packet switch of claim 31 wherein the FPGA includes a processing pipeline, the processing pipeline configured to perform the filter and sort operation and the generate operation in a pipelined fashion as the message data streams through the FPGA.

33. The packet switch of claim 31 wherein the FPGA is further configured to perform a plurality of data consumer-specific search operations on the message data with respect to a plurality of search terms, the search terms being associated with the data consumers to find message data of interest to the data consumers.

34. The packet switch of claim 33 wherein the FPGA, based on the search operations, is further configured to store the message data found to be of interest to the data consumers in a plurality of queues, each queue being associated with a data consumer such that the message data found to be of interest to a particular data consumer is stored in the queue associated with that data consumer.

35. The packet switch of claim 34 wherein the FPGA is further configured to generate the outgoing data consumer-specific data packets from commonly-queued message data.

36. The packet switch of claim 35 wherein the FPGA is further configured to (1) select a queue from which to generate an outgoing data consumer-specific data packet, (2) access packaging parameter data that is associated with the selected queue, and (3) generate an outgoing data consumer-specific data packet from message data in the selected queue in accordance with the accessed packaging parameter data.

37. The packet switch of claim 31 wherein at least a plurality of the received data packets comprise transmission control protocol (TCP) data packets, and wherein the FPGA is further configured to perform a TCP termination on at least a plurality of the received TCP data packets.