Accelerating data transfer in a virtual computer system with tightly coupled TCP connections

• US 8,893,159 B1
• Filed: 09/09/2013
• Issued: 11/18/2014
• Est. Priority Date: 04/01/2008
• Status: Active Grant

• Current Assignee: Alacritech, Inc.
• Original Assignee: Alacritech, Inc.
• Inventors: Starr, Daryl D., Craft, Peter K., Philbrick, Clive M.
• Primary Examiner(s): Sough, H S
• Assistant Examiner(s): Seye, Abdou
• Application Number:
  US14/022,101
• Time in Patent Office: 435 Days
• Field of Search: None
• US Class Current: 719/319
• CPC Class Codes: G06F 2009/45595 Network integration; Enabling network access in virtual machine instances
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/54 Interprogram communication

Abstract

First and second operating systems of a virtual computer system can communicate using respective first and second network protocol stacks, by employing procedures that are specialized for a situation in which a TCP control block of the first stack and a TCP control block of the second stack correspond to the same logical connection. In this case, various TCP requirements can be bypassed by coupling the TCP control blocks, reducing or eliminating data copies and providing other efficiencies.

Claims

1. A method comprising:

running a first operating system and a second operating system on a computer system, the first operating system having a first network protocol stack including a first transmission control protocol (TCP) layer and the second operating system having a second network protocol stack including a second TCP layer;

establishing a logical connection between the first and second TCP layers, including creating, by the first TCP layer, a first TCP control block corresponding to the logical connection, and creating, by the second TCP layer, a second TCP control block corresponding to the logical connection; and

coupling the first TCP control block to the second TCP control block, including executing instructions, by an entity that accesses the first and second TCP control blocks, that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection.

2. The method of claim 1, wherein executing instructions includes referencing the first TCP control block and the second TCP control block within a function call.

3. The method of claim 1, wherein coupling the first TCP control block to the second TCP control block includes associating the first TCP control block with the second TCP control block by a virtual machine manager.

4. The method of claim 1, wherein coupling the first TCP control block to the second TCP control block includes associating the first TCP control block with the second TCP control block by the first operating system.

5. The method of claim 1, wherein coupling the first TCP control block to the second TCP control block includes storing the first TCP control block and the second TCP control block on an interface that is connected to the computer system by an input/output (I/O) channel.

6. The method of claim 1, further comprising recognizing that the first TCP control block and the second TCP control block correspond to the same logical connection.

7. The method of claim 1, further comprising recognizing that the first TCP control block and the second TCP control block are reciprocal to each other.

8. The method of claim 1, further comprising recognizing that the first TCP control block is identified by IP addresses and TCP ports that are a mirror image of the IP addresses and TCP ports of the second TCP control block.

9. The method of claim 1, further comprising flagging the first TCP control block and the second TCP control block as tightly-coupled.

10. The method of claim 1, wherein the first TCP control block points to a location in memory of the second TCP control block, and the second TCP control block points to a location in memory of the first TCP control block.

11. The method of claim 1, further comprising offloading, from the entity to a second entity, the first and second TCPs control blocks as a pair of tightly coupled control blocks.

12. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes transferring data without a corresponding TCP header from a memory source controlled by a process running above the first TCP layer directly to a memory destination controlled by a process running above the second TCP layer.

13. The method of claim 12, wherein transferring the data is performed without the data being copied by a central processing unit (CPU) that runs the first or second operating system, and without the data traversing an I/O channel of the computer.

14. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes transferring, by direct memory access (DMA), data from a memory source controlled by a process running above the first TCP layer to a memory destination controlled by a process running above the second TCP layer.

15. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid transferring window update packets between the first and second TCP layers.

16. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid transferring acknowledgements (ACKs) between the first and second TCP layers.

17. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid segmenting data that is transferred between the first and second TCP layers into packets.

18. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid checksumming data that is transferred between the first and second TCP layers.

19. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid retransmitting data that is transferred between the first and second TCP layers.

20. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid timers.

21. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid round trip time calculations.

22. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes altering the TCP protocol to avoid congestion-control.

23. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes substantially simultaneously referencing the first and second TCP control blocks by the entity.

24. The method of claim 1, wherein executing instructions that are specialized for a situation in which the first and second TCP control blocks correspond to the same logical connection includes substantially simultaneously updating the first and second TCP control blocks by the entity.

25. A method comprising:

running a first operating system and a second operating system on a computer system, the first operating system having a first network protocol stack including a first transmission control protocol (TCP) layer and the second operating system having a second network protocol stack including a second TCP layer;

establishing a logical connection between the first and second TCP layers, including creating, by the first TCP layer, a first TCP control block corresponding to the logical connection, and creating, by the second TCP layer, a second TCP control block corresponding to the logical connection; and

coupling the first TCP control block to the second TCP control block, including executing a function call that references the first and second TCP control blocks.

26. The method of claim 25, wherein executing a function call that references the first and second TCP control blocks includes substantially simultaneously updating the first and second TCP control blocks by the entity.

27. The method of claim 25, wherein coupling the first TCP control block to the second TCP control block includes associating the first TCP control block with the second TCP control block by a virtual machine manager.

28. The method of claim 25, wherein coupling the first TCP control block to the second TCP control block includes associating the first TCP control block with the second TCP control block by the first operating system.

29. The method of claim 25, wherein coupling the first TCP control block to the second TCP control block includes storing the first TCP control block and the second TCP control block on an interface that is connected to the computer system by an input/output (I/O) channel.

30. The method of claim 25, further comprising recognizing that the first TCP control block and the second TCP control block correspond to the same logical connection.

31. The method of claim 25, further comprising recognizing that the first TCP control block and the second TCP control block are reciprocal to each other.

32. The method of claim 25, further comprising recognizing that the first TCP control block is identified by IP addresses and TCP ports that are a mirror image of the IP addresses and TCP ports of the second TCP control block.

33. The method of claim 25, further comprising flagging the first TCP control block and the second TCP control block as tightly-coupled.

34. The method of claim 25, wherein the first TCP control block points to a location in memory of the second TCP control block, and the second TCP control block points to a location in memory of the first TCP control block.

35. The method of claim 25, further comprising offloading, from the entity to a second entity, the first and second TCPs control blocks as a pair of tightly coupled control blocks.

36. The method of claim 25, wherein executing a function call that references the first and second TCP control blocks includes transferring data without a corresponding TCP header from a memory source controlled by a process running above the first TCP layer directly to a memory destination controlled by a process running above the second TCP layer.

37. The method of claim 36, wherein transferring the data is performed without the data being copied by a central processing unit (CPU) that runs the first or second operating system, and without the data traversing an I/O channel of the computer.

38. The method of claim 25, wherein executing a function call that references the first and second TCP control blocks includes transferring, by direct memory access (DMA), data from a memory source controlled by a process running above the first TCP layer to a memory destination controlled by a process running above the second TCP layer.

39. The method of claim 25, further comprising altering the TCP protocol to avoid transferring window update packets between the first and second TCP layers.

40. The method of claim 25, further comprising altering the TCP protocol to avoid transferring acknowledgements (ACKs) between the first and second TCP layers.

41. The method of claim 25, further comprising altering the TCP protocol to avoid segmenting data that is transferred between the first and second TCP layers into packets.

42. The method of claim 25, further comprising altering the TCP protocol to avoid checksumming data that is transferred between the first and second TCP layers.

43. The method of claim 25, further comprising altering the TCP protocol to avoid retransmitting data that is transferred between the first and second TCP layers.

44. The method of claim 25, further comprising altering the TCP protocol to avoid timers.

45. The method of claim 25, further comprising altering the TCP protocol to avoid round trip time calculations.

46. The method of claim 25, further comprising altering the TCP protocol to avoid congestion-control.

47. The method of claim 25, wherein executing a function call that references the first and second TCP control blocks includes substantially simultaneously referencing the first and second TCP control blocks by the entity.

48. A method comprising:

running a first operating system and a second operating system on a computer system, the first operating system having a first network protocol stack including a first transmission control protocol (TCP) layer and the second operating system having a second network protocol stack including a second TCP layer;

establishing a logical connection between the first and second TCP layers, including creating, by the first TCP layer, a first TCP control block corresponding to the logical connection, and creating, by the second TCP layer, a second TCP control block corresponding to the logical connection;

coupling the first TCP control block to the second TCP control block, including referencing the first TCP control block by a set of instructions and referencing the second TCP control block by the set of instructions.