METHOD AND SYSTEM FOR SPLICING REMOTE DIRECT MEMORY ACCESS (RDMA) TRANSACTIONS IN AN RDMA-AWARE SYSTEM

• US 20080301254A1
• Filed: 11/16/2007
• Published: 12/04/2008
• Est. Priority Date: 05/30/2007
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Broadcom Corporation (Broadcom, Inc.)
• Inventors: Bestler, Caitlin
• Granted Patent: US 8,090,790 B2
• Time in Patent Office: Days
• US Class Current: 709/212
• CPC Class Codes: H04L 49/90 Buffering arrangements
• CPC Class Codes: H04L 67/06 specially adapted for file transfer, e.g. file transfer protocol [FTP]
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
• CPC Class Codes: H04L 69/16 Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
• CPC Class Codes: H04L 69/166 IP fragmentation; TCP segmentation

Abstract

Aspects of a system for splicing RDMA transactions in an RDMA system may include a main processor within a main server that may receive read requests from a client device. The main processor may translate a data reference contained in each read request to generate a physical buffer list (PBL). The processor 206 may communicate the PBL to a local processor within a NIC. The local processor may utilize the PBL perform RDMA operations to retrieve data stripes contained in one or more DDP segments received from a plurality of auxiliary servers. The local processor may enable the generation of TCP packets each containing the data stripes, which may then be sent to the client. The retrieval of DDP segments and generation of TCP packets may occur within the NIC without transferring retrieved data stripes via a system bus within the main server.

Claims

1. A method for transferring data in a communications system, the method comprising:

receiving, at a network interface controller (NIC), a read request message that specifies data comprising at least one data stripe,;

generating, within said NIC, at least one address for retrieval of said at least one data stripe;

receiving, at said NIC, at least a portion of said at least one data stripe within at least one direct data placement (DDP) segment based on said generated at least one address; and

transmitting, from said NIC, said received at least a portion of said at least one data strip within at least one packet.

2. The method according to claim 1, wherein said read request comprises a file handle, file offset value and/or file length that identifies said at least one data stripe.

3. The method according to claim 2, comprising selecting an aggregated splicing steering tag (STag) and/or an aggregated tagged offset (TO) value based on said file handle and/or said file offset value.

4. The method according to claim 3, comprising selecting a physical buffer list (PBL) based on said aggregated splicing Stag and/or said aggregated TO value.

5. The method according to claim 4, comprising generating at least one data stripe STag and/or at least one data stripe TO based on said PBL.

6. The method according to claim 5, comprising generating said at least one address based on said generated at least one data stripe STag and/or said generated at least one data stripe TO.

7. The method according to claim 3, comprising generating said packet based on said aggregated splicing STag, at least one aggregated TO index, said file handle, said file offset value and/or said file length.

8. The method according to claim 7, comprising generating said at least one aggregated TO index based on a PBL.

9. The method according to claim 1, comprising transmitting at least one DDP read request message corresponding to said generated at least one address.

10. The method according to claim 9, comprising receiving said at least one DDP segment in response to at least a portion of said transmitted at least one DDP read request message.

11. A machine-readable storage having stored thereon, a computer program having at least one code section for transferring data in a communications system, the at least one code section being executable by a machine for causing the machine to perform steps comprising:

receiving, at a network interface controller (NIC), a read request message that specifies data comprising at least one data stripe;

generating, within said NIC, at least one address for retrieval of said at least one data stripe;

receiving, at said NIC, at least a portion of said at least one data stripe within at least one direct data placement (DDP) segment based on said generated at least one address; and

transmitting, from said NIC, said received at least a portion of said at least one data strip within at least one packet.

12. The machine-readable storage according to claim 11, wherein said read request comprises a file handle, file offset value and/or file length that identifies said at least one data stripe.

13. The machine-readable storage according to claim 12, comprising code for selecting an aggregated splicing steering tag (STag) and/or an aggregated tagged offset (TO) value based on said file handle and/or said file offset value.

14. The machine-readable storage according to claim 13, comprising code for selecting a physical buffer list (PBL) based on said aggregated splicing Stag and/or said aggregated TO value.

15. The machine-readable storage according to claim 14, comprising code for generating at least one data stripe STag and/or at least one data stripe TO based on said PBL.

16. The machine-readable storage according to claim 15, comprising code for generating said at least one address based on said generated at least one data stripe STag and/or said generated at least one data stripe TO.

17. The machine-readable storage according to claim 13, comprising code for generating said packet based on said aggregated splicing STag, at least one aggregated TO index, said file handle, said file offset value and/or said file length.

18. The machine-readable storage according to claim 17, comprising code for generating said at least one aggregated TO index based on a PBL.

19. The machine-readable storage according to claim 11, comprising code for transmitting at least one DDP read request message corresponding to said generated at least one address.

20. The machine-readable storage according to claim 19, comprising code for receiving said at least one DDP segment in response to at least a portion of said transmitted at least one DDP read request message.

21. A system for transferring data in a communications system, the system comprising:

at least one processor that enables reception, at a network interface controller (NIC), of a read request message that specifies data comprising at least one data stripe,;

said at least one processor enables generation, within said NIC, of at least one address for retrieval of said at least one data stripe;

said at least one processor enables reception, at said NIC, of at least a portion of said at least one data stripe within at least one direct data placement (DDP) segment based on said generated at least one address; and

said at least one processor enables transmission, from said NIC, of said received at least a portion of said at least one data strip within at least one packet.

22. The system according to claim 21, wherein said read request comprises a file handle, file offset value and/or file length that identifies said at least one data stripe.

23. The system according to claim 22, wherein said at least one processor enables selection of an aggregated splicing steering tag (STag) and/or an aggregated tagged offset (TO) value based on said file handle and/or said file offset value.

24. The system according to claim 23, wherein said at least one processor enables selection of a physical buffer list (PBL) based on said aggregated splicing Stag and/or said aggregated TO value.

25. The system according to claim 24, wherein said at least one processor enables generation of at least one data stripe STag and/or at least one data stripe TO based on said PBL.

26. The system according to claim 25, wherein said at least one processor enables generation of said at least one address based on said generated at least one data stripe STag and/or said generated at least one data stripe TO.

27. The system according to claim 23, wherein said at least one processor enables generation of said packet based on said aggregated splicing STag, at least one aggregated TO index, said file handle, said file offset value and/or said file length.

28. The system according to claim 27, wherein said at least one processor enables generation of said at least one aggregated TO index based on a PBL.

29. The system according to claim 21, wherein said at least one processor enables transmission of at least one DDP read request message corresponding to said generated at least one address.

30. The system according to claim 29, wherein said at least one processor enables reception of said at least one DDP segment in response to at least a portion of said transmitted at least one DDP read request message.