Method for sharing a transport connection across a multi-processor platform with limited inter-processor communications

• US 7,535,916 B2
• Filed: 02/07/2005
• Issued: 05/19/2009
• Est. Priority Date: 02/07/2005
• Status: Expired due to Fees

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Bryant, Stewart Frederick, Lei, Peter, Stewart, Randall
• Primary Examiner(s): Ryman; Daniel J
• Assistant Examiner(s): Tran; Thinh D
• Application Number:
  US11/053,093
• Publication Number:
  US 20060176904A1
• Time in Patent Office: 1,562 Days
• Field of Search: 370/466, 370/467, 379/229, 379/230, 709/230
• US Class Current: 370/419
• CPC Class Codes: H04L 45/00 Routing or path finding of packets in data switching networks routing or path finding in wireless networks H04W40/00
• CPC Class Codes: H04L 69/26 Special purpose or proprietary protocols or architectures network applications for proprietary or special purpose networking environments H04L67/12

Abstract

A method is disclosed for sharing a transport connection in a host having two or more processors that exchange limited inter-processor communications. Ranges of transport sequence numbers are assigned to the processors of the host, and a range assigned to a particular processor does not overlap with any other range assigned to any other processor. An outbound message is sent over the transport connection from a processor of the host, where the message includes a transport sequence number selected from the range assigned to that processor. An inbound message is received at a designated processor of the host that is configured to receive all inbound messages sent to the host over the transport connection.

Claims

1. A method for sharing a transport connection in a host having two or more processors, the method comprising the computer-implemented steps of:

assigning ranges of transport sequence numbers to the two or more processors, wherein a particular range assigned to a particular processor does not overlap with any other range that is assigned to any other processor;

wherein the two or more processors in the host exchange limited inter-processor communications;

wherein only a designated processor, of the two or more processors, is operable to maintain state information for the transport connection;

sending, over the transport connection, an outbound message from a processor of the two or more processors, wherein the outbound message includes a transport sequence number selected only from the range of the transport sequence numbers assigned to the processor; and

receiving, over the transport connection, an inbound message at the designated processor of the two or more processors, wherein the designated processor is configured to receive all inbound messages sent to the host over the transport connection;

wherein the inbound message includes an acknowledgement of the outbound message.

2. A method as recited in claim 1, further comprising:

generating a set of assigned ranges of the transport sequence numbers; and

assigning a new range of the transport sequence numbers, which is not in the set of assigned ranges, to each of the two or more processors, wherein a particular new range assigned to the particular processor does not overlap with any other range that is assigned to any other processor.

3. A method as recited in claim 2, wherein the steps of generating the set of assigned ranges and assigning the new range of the transport sequence numbers are performed in response to the expiration of a timer.

4. A method as recited in claim 2, wherein the steps of generating the set of assigned ranges and assigning the new range of the transport sequence numbers are performed in response to a request from the particular processor that has used all the transport sequence numbers from the particular range assigned to that particular processor.

5. A method as recited in claim 2, wherein the step of generating the set of assigned ranges comprises polling each processor of the two or more processors to determine the ranges of the transport sequence numbers that have been assigned to said each processor.

6. A method as recited in claim 2, wherein the step of generating the set of assigned ranges comprises checking a table in memory to determine the ranges of the transport sequence numbers that have been assigned.

7. A method as recited in claim 1, further comprising the step of sending to a peer host a forward cumulative transport sequence number, wherein the peer host skips any one or more missing messages including any one or more transport sequence numbers that are smaller than or equal to the forward cumulative transport sequence number.

8. A method as recited in claim 1, wherein the step of assigning the ranges of the transport sequence numbers is performed by making use of an inter-processor communication mechanism.

9. A method as recited in claim 1, wherein the transport connection is a Datagram Congestion Control Protocol (DCCP) connection.

10. A method for sharing a Stream Control Transmission Protocol (SCTP) association across a host having two or more processors, the method comprising the computer-implemented steps of:

establishing an SCTP association between the host and a peer host, wherein a designated processor of the two or more processors is configured to receive all inbound messages sent over the SCTP association by the peer host;

wherein the inbound messages include acknowledgements of all corresponding outbound messages that are sent to the peer host over the SCTP association by the two or more processors;

wherein the two or more processors in the host exchange limited inter-processor communications;

wherein only the designated processor is operable to maintain state information for the SCTP association;

assigning a range of transport sequence numbers to each of the two or more processors, wherein a particular range assigned to a particular processor does not overlap with any other range that has been assigned to any other processor; and

sending an outbound message from a processor of the two or more processors to the peer host over the SCTP association, wherein the outbound message includes a verification tag of the SCTP association and a transport sequence number selected from the range of the transport sequence numbers assigned to the processor.

11. A method as recited in claim 10, further comprising:

generating a set of assigned ranges of the transport sequence numbers;

assigning a new range of the transport sequence numbers, which is not in the set of assigned ranges, to each of the two or more processors, wherein a particular new range assigned to the particular processor does not overlap with any other range that is assigned to any other processor; and

sending a forward cumulative transport sequence number to the peer host over the SCTP association;

wherein sending the forward cumulative transport sequence number to the peer host comprises causing the peer host to skip past any one or more missing messages that include any one or more transport sequence numbers that are smaller than or equal to the forward cumulative transport sequence number.

12. A method as recited in claim 11, wherein the steps of generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to the expiration of a timer.

13. A method as recited in claim 11, wherein the steps of generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to a request from a particular processor that has used all the transport sequence numbers from the particular range assigned to that particular processor.

14. A method as recited in claim 11, wherein the steps of generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to failing to receive, from the peer host, acknowledgements of a preset number of messages.

15. A method as recited in claim 11, wherein the steps of generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to receiving, from the peer host, acknowledgements of a preset number of messages.

16. A method as recited in claim 11, wherein the step of generating the set of assigned ranges comprises polling each processor of the two or more processors to determine the ranges of the transport sequence numbers that have been assigned to said each processor.

17. A method as recited in claim 11, wherein the step of generating the set of assigned ranges comprises checking a table in memory to determine the ranges of the transport sequence numbers that have been assigned to the two or more processors.

18. A method as recited in claim 10, further comprising assigning a different stream of the SCTP association to each of the two or more processors, wherein each processor of the two or more processors sends messages only over the stream assigned to said each processor.

19. A method as recited in claim 10, wherein:

the host comprises a router; and

the peer host comprises a collector that is configured to collect network management information.

20. A method as recited in claim 10, wherein the SCTP association is established according to a Partial Reliability Stream Control Transmission Protocol (PR-SCTP) standard.

21. An apparatus comprising:

a memory;

two or more processors operatively coupled to the memory, wherein;

the two or more processors are operable to exchange limited inter-processor communications;

the two or more processors are operable to share a transport connection;

each of the two or more processors is operatively connectable to a network; and

only a designated processor, of the two or more processors, is operable to maintain state information for the transport connection;

one or more stored sequences of instructions which, when executed, cause;

assigning ranges of transport sequence numbers to the two or more processors, wherein a particular range assigned to a particular processor does not overlap with any other range that is assigned to any other processor;

sending, over the transport connection, an outbound message from a processor of the two or more processors, wherein the outbound message includes a transport sequence number selected only from the range of the transport sequence numbers assigned to the processor; and

receiving, over the transport connection, an inbound message at the designated processor of the two or more processors, wherein the designated processor is configured to receive all inbound messages that are sent over the transport connection;

wherein the inbound message includes an acknowledgement of the outbound message.

22. An apparatus as recited in claim 21, wherein the apparatus comprises a router.

23. The apparatus of claim 21, wherein the one or more stored sequences of instructions further comprise instructions which, when executed, cause:

generating a set of assigned ranges of the transport sequence numbers; and

assigning a new range of the transport sequence numbers, which is not in the set of assigned ranges, to each of the two or more processors, wherein a particular new range assigned to the particular processor does not overlap with any other range that is assigned to any other processor.

24. The apparatus of claim 23, wherein the instructions that cause generating the set of assigned ranges and assigning the new range of the transport sequence numbers are performed in response to one or more of:

the expiration of a timer; and

a request from a particular processor that has used all the transport sequence numbers from the particular range assigned to that particular processor.

25. The apparatus of claim 23, wherein the instructions that cause generating the set of assigned ranges comprise instructions which, when executed, cause one or more of:

polling each processor of the two or more processors to determine the ranges of the transport sequence numbers that have been assigned to said each processor; and

checking a table in the memory to determine the ranges of the transport sequence numbers that have already been assigned.

26. The apparatus of claim 21, wherein the one or more stored sequences of instructions further comprise instructions which, when executed, cause sending to a peer host a forward cumulative transport sequence number, wherein the peer host skips any one or more missing messages including any one or more transport sequence numbers that are smaller than or equal to the forward cumulative transport sequence number.

27. A computer-readable storage medium storing one or more sequences of instructions for sharing a transport connection across a host having two or more processors, which instructions, when executed by one or more processors, cause the one or more processors to carry out the steps of:

assigning ranges of transport sequence numbers to the two or more processors, wherein a particular range assigned to a particular processor does not overlap with any other range that is assigned to any other processor;

wherein the two or more processors in the host exchange limited inter-processor communications;

wherein only a designated processor, of the two or more processors, is operable to maintain state information for the transport connection;

sending, over the transport connection, an outbound message from a processor of the two or more processors, wherein the outbound message includes a transport sequence number selected only from the range of the transport sequence numbers assigned to the processor; and

receiving, over the transport connection, an inbound message at the designated processor of the two or more processors, wherein the designated processor is configured to receive all inbound messages sent to the host over the transport connection;

wherein the inbound message includes an acknowledgement of the outbound message.

28. A computer-readable storage medium as recited in claim 27, further comprising instructions which, when executed by the one or more processors, cause the one or more processors to carry out the steps of:

generating a set of assigned ranges of the transport sequence numbers; and

assigning a new range of the transport sequence numbers, which is not in the set of assigned ranges, to each of the two or more processors, wherein a particular new range assigned to the particular processor does not overlap with any other range that is assigned to any other processor.

29. A computer-readable storage medium as recited in claim 27, further comprising instructions which, when executed by the one or more processors, cause the one or more processors to carry out the step of sending to a peer host a forward cumulative transport sequence number, wherein the peer host skips any one or more missing messages including any one or more transport sequence numbers that are smaller than or equal to the forward cumulative transport sequence number.

30. A computer-readable storage medium as recited in claim 27, wherein the transport connection is a Datagram Congestion Control Protocol (DCCP) connection.

31. A computer-readable storage medium storing one or more sequences of instructions for sharing a Stream Control Transmission Protocol (SCTP) association across a host having two or more processors, which instructions, when executed by one or more processors, cause the one or more processors to carry out the steps of:

establishing the SCTP association between the host and a peer host, wherein a designated processor of the two or more processors is configured to receive all inbound messages sent over the SCTP association by the peer host;

wherein the inbound messages include acknowledgements of all corresponding outbound messages that are sent to the peer host over the SCTP association by the two or more processors;

wherein the two or more processors in the host exchange limited inter-processor communications;

wherein only the designated processor is operable to maintain state information for the SCTP association;

assigning a range of transport sequence numbers to each of the two or more processors, wherein a particular range assigned to a particular processor does not overlap with any other range that has been assigned to any other processor; and

sending an outbound message from a processor of the two or more processors to the peer host over the SCTP association, wherein the outbound message includes a verification tag of the SCTP association and a transport sequence number selected from the range of the transport sequence numbers assigned to the processor.

32. A computer-readable storage medium as recited in claim 31, further comprising instructions which, when executed by the one or more processors, cause the one or more processors to carry out the steps of:

generating a set of assigned ranges of the transport sequence numbers;

assigning a new range of the transport sequence numbers, which is not in the set of assigned ranges, to each of the two or more processors, wherein a particular new range assigned to the particular processor does not overlap with any other range that is assigned to any other processor; and

sending a forward cumulative transport sequence number to the peer host over the SCTP association;

wherein sending the forward cumulative transport sequence number to the peer host comprises causing the peer host to skip past any one or more missing messages that include any one or more transport sequence numbers that are smaller than or equal to the forward cumulative transport sequence number.

33. A computer-readable storage medium as recited in claim 32, wherein the instructions for carrying out the steps of generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number further comprise instructions for carrying out the steps in response to a request from a particular processor that has used all the transport sequence numbers from the particular range assigned to that particular processor.

34. A computer-readable storage medium as recited in claim 32, wherein the instructions for carrying out the steps of generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number further comprise instructions for carrying out the steps in response to receiving, from the peer host, acknowledgements of a preset number of messages.

35. A computer-readable storage medium as recited in claim 31, further comprising instructions which, when executed by the one or more processors, cause the one or more processors to carry out the step of assigning a different stream of the SCTP association to each of the two or more processors, wherein each processor of the two or more processors sends messages only over the stream assigned to said each processor.

36. A computer-readable storage medium as recited in claim 31, wherein the SCTP association is established according to a Partial Reliability Stream Control Transmission Protocol (PR-SCTP) standard.

37. An apparatus comprising:

a memory;

two or more processors operatively coupled to the memory, wherein;

the two or more processors are operable to share a Stream Control Transmission Protocol (SCTP) association;

the two or more processors are operable to exchange limited inter-processor communications;

each of the two or more processors is operatively connectable to a network; and

only a designated processor, of the two or more processors, is operable to maintain state information for the SCTP association;

one or more stored sequences of instructions which, when executed, cause;

establishing the SCTP association to a peer host, wherein the designated processor of the two or more processors is configured to receive all inbound messages sent over the SCTP association by the peer host;

wherein the inbound messages include acknowledgements of all corresponding outbound messages that are sent to the peer host over the SCTP association by the two or more processors;

assigning a range of transport sequence numbers to each of the two or more processors, wherein a particular range assigned to a particular processor does not overlap with any other range that has been assigned to any other processor; and

sending an outbound message from a processor of the two or more processors to the peer host over the SCTP association, wherein the outbound message includes a verification tag of the SCTP association and a transport sequence number selected from the range of the transport sequence numbers assigned to the processor.

38. The apparatus of claim 37, wherein the one or more stored sequences of instructions further comprise instructions which, when executed, cause:

generating a set of assigned ranges of the transport sequence numbers;

assigning a new range of the transport sequence numbers, which is not in the set of assigned ranges, to each of the two or more processors, wherein a particular new range assigned to the particular processor does not overlap with any other range that is assigned to any other processor; and

sending a forward cumulative transport sequence number to the peer host over the SCTP association;

wherein sending the forward cumulative transport sequence number to the peer host comprises causing the peer host to skip past any one or more missing messages that include any one or more transport sequence numbers that are smaller than or equal to the forward cumulative transport sequence number.

39. The apparatus of claim 38, wherein the instructions that cause generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to one or more of:

the expiration of a timer; and

a request from a particular processor that has used all the transport sequence numbers from the particular range assigned to that particular processor.

40. The apparatus of claim 38, wherein the instructions that cause generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to failing to receive, from the peer host, acknowledgements of a preset number of messages.

41. The apparatus of claim 38, wherein the instructions that cause generating the set of assigned ranges, assigning the new range of the transport sequence numbers, and sending the forward cumulative transport sequence number are performed in response to receiving, from the peer host, acknowledgements of a preset number of messages.

42. The apparatus of claim 38, wherein the instructions that cause generating the set of assigned ranges comprise instructions which, when executed, cause one or more of:

polling each processor of the two or more processors to determine the ranges of the transport sequence numbers that have been assigned to said each processor; and

checking a table in the memory to determine the ranges of the transport sequence numbers that have been assigned to the two or more processors.

43. The apparatus of claim 37, wherein the one or more stored sequences of instructions further comprise instructions which, when executed, cause assigning a different stream of the SCTP association to each of the two or more processors, wherein each processor of the two or more processors sends messages only over the stream assigned to said each processor.