CN100472450C - System and method for network interfacing in a multiple network environment - Google Patents

System and method for network interfacing in a multiple network environment Download PDF

Info

Publication number
CN100472450C
CN100472450C CNB038201275A CN03820127A CN100472450C CN 100472450 C CN100472450 C CN 100472450C CN B038201275 A CNB038201275 A CN B038201275A CN 03820127 A CN03820127 A CN 03820127A CN 100472450 C CN100472450 C CN 100472450C
Authority
CN
China
Prior art keywords
processor
service
networks
network
different type
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CNB038201275A
Other languages
Chinese (zh)
Other versions
CN1757016A (en
Inventor
尤里·埃尔朱
弗兰克·凡
史蒂夫·林赛
斯科特·S·麦克丹尼尔
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Broadcom Corp
Zyray Wireless Inc
Original Assignee
Zyray Wireless Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US10/230,643 external-priority patent/US7295555B2/en
Application filed by Zyray Wireless Inc filed Critical Zyray Wireless Inc
Publication of CN1757016A publication Critical patent/CN1757016A/en
Application granted granted Critical
Publication of CN100472450C publication Critical patent/CN100472450C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Abstract

Systems and methods that identify the Upper Layer Protocol (ULP) message boundaries are provided. In one example, a method that identifies ULP message boundaries is provided. The method may include one or more of the following steps: attaching a framing header of a frame to a data payload to form a packet, the framing header being placed immediately after the byte stream transport protocol header, the framing header comprising a length field comprising a length of a framing protocol data unit (PDU); and inserting a marker in the packet, the marker pointing backwards to the framing header and being inserted at a preset interval.

Description

Be used for the system and method that network connects in the multi-network environment
Cross reference/combination by reference to related application
The application's reference, the name that requires right of priority and requirement to benefit to submit on June 10th, 2003 are called the U.S. Provisional Patent Application of " being used for the system and method (System and Method forNetwork Interfacing in a Multiple Network Environment) that network connects in multi-network environment ", its sequence number is 60/477,279; The name of submitting on June 11st, 2003 is called the U.S. Provisional Patent Application of " being used for the system and method (System and Method for Network Interfacing) that network connects ", and its sequence number is 60/478,106; The name of submitting on September 6th, 2002 is called the U.S. Provisional Patent Application of " system and method (System and Method for TCP/IP Offload) that is used for the TCP/IP unloading ", and its sequence number is 60/408,617; The name of submitting on August 30th, 2002 is called the U.S. Provisional Patent Application of " system and method (System and Method for TCP Offload) that is used for the TCP unloading ", and its sequence number is 60/407,165; The name of submitting on March 30th, 2003 is called the U.S. Provisional Patent Application of " system and method (System and Method for TCP Offload) that is used for the TCP unloading ", and its sequence number is 60/456,265; The name of submitting on March 20th, 2003 is called the U.S. Provisional Patent Application of " system and method (System and Method for Handling Out-of-Order Frames) that is used to handle unordered frame ", and its sequence number is 60/456,260; The name of submitting on September 11st, 2002 is called the U.S. Provisional Patent Application of " system and method (System and Method for TCPOffloading and Uploading) that is used for the TCP unloading and loads ", and its sequence number is 60/410,022; The name of submitting on November 18th, 2002 is called the U.S. Patent application of " system and method (System and Method for TCP Offloading and Uploading) that is used for the TCP unloading and loads ", and its sequence number is 10/298,817; The name of submitting on September 17th, 2002 is called the U.S. Provisional Patent Application of " system and method (System and Method for Handling Partially ProcessedFrames) that is used for the frame of administrative section processing ", its sequence number is 60/411,294; The name of submitting on November 21st, 2002 is called the U.S. Patent application of " system and method (System andMethod for Handling Frames in Multiple Stack Envirments) that is used for handling the frame of multiple stack environment ", its sequence number is 10/302,474; The name of submitting on September 4th, 2002 is called the U.S. Provisional Patent Application of " system and method (System and Method for Fault Tolerant TCP Offload) that is used for fault-tolerant TCP unloading ", and its sequence number is 60/408,207; The name of submitting on January 6th, 2003 is called the U.S. Patent application of " system and method (System and Method for Fault Tolerant TCPOffload) that is used for fault-tolerant TCP unloading ", and its sequence number is 10/337,029; The name of submitting on August 23rd, 2002 is called the U.S. Provisional Patent Application of " using the long-range direct memory access (DMA) (Remote Direct Memory Access over TCP/IP using Generic Buffers forNon-Posting TCP) on the TCP/IP of the universal buffer that does not upgrade TCP ", its sequence number is 60/405,539; The name of submitting on July 26th, 2002 is called the U.S. Provisional Patent Application of " be used for storehouse and directly visit the double T CP/IP stack connection management of (WSD) (Dual TCP/IP Stacks Connection Management for Winsock Direct (WSD)) ", its sequence number is 60/398,663; The name of submitting on Dec 18th, 2002 is called the U.S. Provisional Patent Application of " system and method (System and Method for HandlingMultiple Stack Environment) that is used to handle the multiple stack environment ", its sequence number is 60/434,503; The name of submitting on January 6th, 2003 is called the U.S. Patent application of " system and method (System andMethod for Handling Multiple Stack Environment) that is used to manage the multiple stack environment ", and its sequence number is 10/336,983; The name of submitting on August 14th, 2002 is called the U.S. Provisional Patent Application of " an emission RDMA (One Shot RDMAHaving Only a 2 Bit State) that 2 bit status are only arranged ", and its sequence number is 60/403,817; The name of submitting on August 19th, 2002 is called the U.S. Provisional Patent Application of " the optimization RDMA (Optimizing RDMA for Storage Provisioning) that is used to store application ", and its sequence number is 60/404,709; The name of submitting on October 18th, 2002 is called the U.S. Patent application of " system and method (System and Method for Statistical Provisioning) that is used to add up configuration ", and its sequence number is 60/419,354; The name of submitting on October 24th, 2002 is called the U.S. Patent application of " system and method (System and Method for Statistical Provisioning) that is used to add up configuration ", and its sequence number is 60/420,901; The name of submitting on January 14th, 2003 is called the U.S. Patent application of " system and method (System and Method for StatisticalProvisioning) that is used to add up configuration ", and its sequence number is 60/439,951; The name of submitting on January 24th, 2003 is called the U.S. Patent application of " system and method (System and Method forStatistical Provisioning) that is used to add up configuration ", and its sequence number is 60/442,360; The name of submitting on November 12nd, 2002 is called the U.S. Provisional Patent Application of " the stored in association management (JointMemory Management for User Space and Storage) that is used for user's space and storage ", and its sequence number is 60/425,959; The name of submitting on March 20th, 2003 is called the U.S. Provisional Patent Application of " self-described host-host protocol section (Self-Describing Transport Protocol Segments) ", and its sequence number is 60/456,266; The name of submitting on January 2nd, 2003 is called the U.S. Provisional Patent Application of " header alignment and complete PDU (Header Alignment and Complete PDU) ", and its sequence number is 60/437,887; The name of submitting on March 20th, 2003 is called the U.S. Provisional Patent Application of " system and method (System and Method for Handling Transport Protocol Segments) that is used to handle the host-host protocol section ", its sequence number is 60/456,322; The name of submitting on September 29th, 2002 is called the U.S. Patent application of " system and method (System and Method forIdentifying Upper Layer Protocol Message Boundaries) that is used to identify the upper-layer protocol message boundaries ", its sequence number is 10/230,643.
U.S. Patent application cited above intactly is combined in herein by reference.
Technical field
Some embodiment of the present invention relates to network interface.More specifically, some embodiment of the present invention relates to and is used for method of network connection and system in the multi-network environment.
Background technology
Fig. 1 has shown the server 100 that is applicable to the Network of handling five types.First type Network is typical Network (professional or other 3 layers of (L3) technology in transmission over networks low volume data and control information of public ethernet network that for example comprise Internet Protocol (IP).First type Network is handled by first network traffic system, and first network traffic system comprises ethernet connector 110; 2 layers of (L2) network interface unit (NIC) device 120 that comprises L2 NIC130; , Peripheral Component Interface (PCI) bridge 140; L2 NIC driver 150; Full feature software transmission control protocol (TCP) stack 160; Socket service switch 170 and socket service 180.Full feature software transmission control protocol (TCP) stack 160 support socket services and other service.
It is professional that second type Network is that TCP quickens, such as, operate in the TCP at IP top.This agreement is used for transmitting big data on conventional Ethernet.The TCP part of server 100 unloading Networks is used to move non-networking task thereby discharge server resource.Second type Network is handled by second network traffic system that comprises TCP offload engine (TOE), and described TOE quickens the TCP business.Second network traffic system comprises ethernet connector 190; 4 layers of (L4) offload adapter arrangement 200 that comprise L2 NIC 210 and TCP processor 220; PCI bridge 140; L4 driver 230; Socket service switch 170 and socket service 180.Generally quicken business service for TCP by socket service 180.
The Network of the third type is a storage service.Conventional storage system uses small computer system interface (SCSI) or fibre channel technology so that server 100 is connected to memory disk.This two kinds of technology sharing of common software interfaces or service, i.e. SCSI mini-port.Current, developed agreement, it provides the SCSI that operates on the TCP/IP network business.Current agreement has been eliminated the needs that SCSI is connected with fibre channel networks, thus make storage service can with the networking employed identical network on (for example Ethernet) operation.By comprising that the 3rd network traffic system of adapter handles the Network of the third type, described adapter realizes that current agreement also provides the service of SCSI mini-port.The 3rd network traffic system comprises ethernet connector 240; Storage host bus adapter (HBA) device 250 that comprises L2 NIC 260, TCP processor 270 and internet SCSI (iSCSI) processor 280; PCI bridge 140; SCSI driver 290 and SCSI mini-port service 300.
The 4th type Network is interprocess communication (IPC) business.This network can move with high-speed traffic program on different server, and has low-down expense.IPC network and use together such as Distributed Application, database server and file server.For example when needed rated output exceeds the ability of particular server or when multiple server is used for super reliable operation, can use the IPC network.Provide this kind service by long-range direct memory access (DMA) (RDMA) interface (for example storehouse is directly visited (Winsock Direct)) that directly is connected with application.The 4th network traffic system is handled the 4th type Network, and described the 4th network traffic system comprises adapter, and this adapter provides the service as special-purpose, privately-owned network (for example unlimited broadband product).The 4th network traffic system comprises proprietary network interface 310; The RDMA NIC device 320 that comprises L2 NIC330, L4 processor and RDMA processor 340; PCI bridge 140; RDMA driver 350 and RDMA service 360 (for example Winsock Direct).
The 5th type Network is and the unknown management entity of any kind of operating system (OS) or device-dependent any business.The state of these entities or monitoring of tools server 100 and transmit with network on the state information relevant with statistical value.The 5th network traffic system is handled the 5th type Network, and the 5th network traffic system comprises ethernet connector 370; Server admin agency 380 and keyboard/video/mouse service 390.The 5th network traffic system provides keyboard, video and mouse hardware service for server 100, so that these interfaces can be redirected to the central server management system on network.
The 5th network traffic system that server 100 is supported is used sizable space in the server, and generally is very expensive.Five types combination of network is hindered in some front portions.For example many operating systems adhere to that each connector should have the driver of himself.Correspondingly, each of five kinds of network traffic system all has data and the control path of oneself.In addition, the use of proprietary network interface makes integrated minimizing possibility.Like this, the redundancy and the ineffectivity that have kept many hardware and softwares.
By such system and in remainder of the present invention, proposed with reference to the accompanying drawings more of the present invention aspect comparison, for the person skilled of this area, other limitation and shortcoming conventional and classic method are conspicuous.
Summary of the invention
In some embodiment of the system and method that network interface is provided, can find of the present invention aspect.In one embodiment, the present invention can provide server.This server can comprise, for example network connector, processor, Peripheral Component Interface (PCI) bridge and unified driver.Processor can be connected to network connector and PCI bridge.Processor is applicable to, for example handles the Network of number of different types.Unified driver can be connected to the PCI bridge, and this unification driver is applicable to the driver that is associated with the Network of number of different types is provided.
In another embodiment, the invention provides and be used for method of network connection.This method can comprise that for example following is one or more: manage a plurality of networks of different type business via 2 layers of (L2) connector; Handle the networks of different type business in the single-chip; And determine which access software service in different types of industry Network via the forms data path.
And in another embodiment, the present invention can be provided for method of network connection.This method comprises that for example following is one or more: handle a plurality of networks of different type business via single ethernet connector; Use L2 processor, 3 layers of (L3) processor, 4 layers of (L4) processor and a plurality of networks of different type business of upper-layer protocol (ULP) processor processing; And provide unified data and control path.
The present invention's these or other advantage, aspect and novel features, and the details of illustrated embodiment of the present invention will be understood from following description and accompanying drawing more fully.
Description of drawings
Fig. 1 illustrates block representation, and the embodiment of server is described;
Fig. 2 illustrates block representation, and an embodiment of the server according to the present invention is described;
Fig. 3 illustrates block representation, and an embodiment of the server according to the present invention is described.
Embodiment
Can in the system and method that network interface for example is provided, find more of the present invention aspect.Can provide system and method according to some embodiments of the present invention in conjunction with network function.For example according to one embodiment of present invention, common network adapter, storage adapter, interprocess communication (IPC) adapter and management adapter combination can be become single equipment.For example, can obtain the abundant saving on cost and the space by time division multiplex sharing module resource or by the fixed resource between the dynamic assignment different network type.Exploitation is shared piece so that the feature that is applicable to one or more agreements (for example function) to be provided.Shared piece also can hold not by the employed special service of institute's protocols having.
Fig. 2 illustrates block representation, and an embodiment of the server 400 according to the present invention is described.Server 400 can comprise for example ethernet connector 410 and server shell 420.The present invention also considers and uses one or more ethernet connectors 410.For example use additional ethernet connector 410 with the performance that enhancing is provided, fault-tolerant or cooperation.Server 400 is applicable to handling a plurality of heterogeneous networks via one or more ethernet connectors 410.As shown in the figure, according to one embodiment of the invention, server 400 can be handled five kinds of networks of different type business.Yet the present invention considers that also processing is greater or less than five kinds of networks of different type business.
Server 400 accessible first type Networks can be common network business (for example adopt the Internet (IP) technology for example or other 3 layers of (L3) technology and in the ethernet network business of transmission over networks low volume data and control information) for example.First network traffic system can be handled first type Network, and first network traffic system can comprise for example 410,2 layers of (L2) network interface unit of the Internet controller (NIC) 430, Peripheral Component Interface (PCI) 440, unified driver 450, software transmission control protocol (TCP) processor 460, socket service switch 470 and socket service 480.Ethernet connector 410 can be connected to L2 NIC430, described L2NIC 430 is sequentially connected to PCI bridge 440.PCI bridge 440 can be connected to unified driver 450, described unified driver 450 is sequentially connected to software TCP processor 460.TCP processor 460 can be connected to socket service switch 470, described socket service switch 470 is sequentially connected to socket service 480.Software TCP processor 460 can be supported for example socket service and other type of service.
Server 400 accessible second type Networks can be that for example TCP quickens professional (for example operating in the TCP at IP top).TCP on the IP is used in and transmits mass data on the Ethernet.Server 400 can unload the TCP part of Network, thereby discharges server resource, is used to move non-networking task.Second network traffic system can be handled second type Network, and second network traffic system comprises the TCP offload engine (TOE) that for example can quicken the TCP business.Second network traffic system comprises for example ethernet connector 410, L2 NIC430, TCP processor 490, PCI bridge 440, unified driver 450, socket service switch 470 and socket service 480.Ethernet connector 410 is connected to L2NIC430, and described L2 NIC 430 is sequentially connected to TCP processor 490.TCP processor 490 is connected to the PCI bridge, and described PCI bridge is linked unified driver 450 in proper order.Consolidator 450 is connected to socket service switch 470, and described socket service switch 470 is sequentially connected to socket service 480.Can be that TCP quickens business service by socket service 480 or other type of service.
The Network of server 400 accessible the third types can be a storage service for example.The Network of the third type comprises the agreement (for example internet SCSI (iSCSI)) that small computer system interface (SCSI) for example is provided on the TCP/IP network.By using iSCSI, can avoid using proprietary adapters, and storage service can move on by some or all networks shared in the networks of different type business.The 3rd network traffic system can be handled the Network of the third type, and the 3rd network traffic system comprises for example ethernet connector 410, L2 NIC430, TCP processor 490, the long-range direct memory access (DMA) of iSCSI/ (RDMA) processor 500, PCI bridge 440, unified driver 450 and SCSI mini-port service 510.Ethernet connector 410 can be connected to L2 NIC 430, described NIC430 is sequentially connected to TCP processor 490.TCP processor 490 can be connected to iSCSI/RDMA processor 500, described iSCSI/RDMA processor 500 is sequentially connected to PCI bridge 440.PCI bridge 440 can be connected to unified driver 450, described unified driver 450 is sequentially connected to SCSI mini-port service 510.
Server 400 manageable the 4th type Networks can be IPC business for example.The IPC network can make program run on different server so that high-speed traffic, and do not have sizable expense.The IPC network can use with for example Distributed Application, database server and file server.For example when essential rated output exceeds the capacity of particular server or when a plurality of servers are used for super reliable operation, can use the IPC network.(for example Winsock Direct that directly is connected with application) can provide such service by the RDMA interface.The 4th network traffic system can be handled the 4th type Network, and the 4th network traffic system can comprise for example ethernet connector 410, L2 NIC 430, TCP processor 490, iSCSI/RDMA processor 500, PCI bridge 440, unified driver 450 and RDMA service 520 (for example Winsock Direct).Ethernet connector 410 is connected to L2 NIC430, and described L2 NIC430 is sequentially connected to TCP processor 490.TCP processor 490 is connected to iSCSI/RDMA processor 500, and described iSCSI/RDMA processor 500 is sequentially connected to PCI bridge 440.PCI bridge 440 is connected to unified driver 450, and described unified driver 450 is sequentially connected to RDMA service 520.
Server 400 accessible the 5th type Networks can be for example with the unknown management entity of operating system (OS) or device-dependent any business of any type.But the state of these entities or equipment monitor server 400, and can be on network the transmission information relevant with state and statistical value.The 5th network traffic system can be handled the 5th type Network, and the 5th network traffic system can comprise for example ethernet connector 410, L2 NIC 430, server admin agency 530 and keyboard/video/mouse service 540.The 5th network traffic system can be server 400 keyboard, video and mouse hardware service is provided, so that these interfaces can be redirected to central server management system (not shown) on network.Ethernet connector 410 can be connected to L2 NIC430, described L2 NIC430 is sequentially connected to server admin agency 530.Server admin can be acted on behalf of 530 and be connected to keyboard/video/mouse service 540.
The present invention considers and adopts different integrated levels.For example according to one embodiment of present invention, single integrated chip 550 can comprise one or more in for example following: L2 NIC430, TCP processor 490 and iSCSI/RDMA processor 500.According to another embodiment of the present invention, software 560 for example can provide one or more in following: TCP processor 460, socket service switch 470, socket service 480, unified driver 450, SCSI mini-port service 510, RDMA service 520 and keyboard/video/mouse service 540.
Fig. 3 illustrates block representation, and the server 400 that has some integrated components according to of the present invention is described.In according to one embodiment of present invention, server shell 420 holds single integrated chip 550, server admin agency 530, PCI bridge 440 and software 560.Single integrated chip 550 is connected to ethernet connector 410, PCI bridge 440 and service management agent 530.Can be with each is connected to software 560 in PCI bridge 440 and the service management agent 530.Like this, single integrated chip 550 can be handled for example five types Networks by single ethernet connector 410.Single integrated chip 550 or PCI bridge 440 are determined which the addressable software 560 in five types the Network, and software 560 comprises unified driver 450 and various service 480,510,520 and 540.Can realize visit via many different technology (comprising for example fixed resource between the time division multiplex and dynamic assignment different network type) to software 560.
Comprise one or more advantages as described below according to some embodiments of the present invention.
Can adopt single ethernet connector 410 according to some embodiments of the present invention.Some embodiment can overcome the difficulty that each connector that the certain operations system adheres to being used for every type business should have the different driving program.In addition, other embodiment can avoid using 3 layers (L3) exchange (switch), need 3 layers (L3) exchange to combine with the particular ip address that will connect and handle specific transactions, this is because must the incoming traffic classification also only be sent to special adapter with incoming traffic by the IP address.
Provide unified data path and control path according to some embodiments of the present invention.This unified method can provide a large amount of savings in cost and space by integrated different parts.
Can share TCP stack between the networks of different type operation system according to some embodiments of the present invention.Cost savings are to produce by eliminating redundant logic and code.
Can share cache memory package according to some embodiments of the present invention.Because network traffic system sharing of common Ethernet connects, network traffic system can be shared reception (RX) and emission (TX) memory buffer.
Can share direct memory access (DMA) (DMA) engine and buffering technology according to some embodiments of the present invention.Some network traffic system and agreement are shared buffering strategy, thereby can share the logic that is used to shine upon.In addition, because the DMA business can use single Ethernet to connect, buffering strategy can be shared identical DMA structure.
Have similar NIC-driver and driver-NIC strategy according to some embodiments of the present invention.By using the communal technique of two kinds of direction indications, can on independent implementation, save cost.
Use single IP address according to some embodiments of the present invention.By many networks and function are combined among single NIC, can adopt single IP address to come to provide service for all these networks and function.This can fully reduce the quantity of the IP address of using in the server system of complexity, also can simplify the management and the configuration of this system.
The centralized control and/or the dynamic assignment of JA(junction ambient) data can be provided according to some embodiments of the present invention.The centralized control of the JA(junction ambient) between different agreement can fully reduce the storage space of use and can realize the storage of JA(junction ambient) in on-chip memory is realized.On-chip memory is realized eliminating the pins/power complexity that for example is associated with external memory storage.
Though described the present invention, those skilled in the art should understand that and to make various changes and of equal value the replacement without departing from the scope of the invention with reference to some embodiment.In addition, under the prerequisite that does not break away from this scope, can make some modifications so that specific environment or material are applicable to teaching of the present invention.Therefore, the present invention does not plan to be confined to disclosed specific embodiment, yet the present invention will comprise all embodiment that fall into the appended claims protection domain.
Correspondingly, the combination of available hardware, software or hardware and software realizes the present invention.Can be in a computer system be distributed on several interconnected computer systems in a concentrated manner or at different elements and realize the present invention in the mode that distributes.Be applicable to that computer system or other device of any kind of of described method all are suitable in execution herein.The combination of typical hardware and software is the general-purpose computing system with computer program, and when loading and carrying out this computer program, the control computer system is so that it carries out method described herein.
Also can be with in the embeddeding computer program product of the present invention, this computer program comprises all features that can carry out methods described herein, can carry out these methods when loading this computer program in computer system.Computer program in the content of the present invention can be the expression formula with the instruction set of any language, code or symbol, this instruction set be used to make have information processing capability system directly, after any one of following process or after following two processes, carry out specific function: a) be converted to another kind of language, code or symbol; B) reproduce with the different material form.
Though described the present invention, those skilled in the art should understand that and to make various changes and of equal value the replacement without departing from the scope of the invention with reference to some embodiment.In addition, under the prerequisite that does not break away from this scope, can make some modifications so that specific environment or material are applicable to teaching of the present invention.Therefore, the present invention does not plan to be confined to disclosed specific embodiment, yet the present invention will comprise all embodiment that fall into the appended claims protection domain.

Claims (28)

1. a server is characterized in that, comprises network connector, processor, network device interface bridge and unified driver, wherein:
Described processor is connected to described network connector, and described processor is applicable to a plurality of networks of different type business of handling, and makes different type network business from network connector via the unified driver of forms data path visit;
Described processor is received in described Peripheral Component Interface bridging; And
Described unified driver is connected to described Peripheral Component Interface bridge, and described unified driver is applicable to provides the driver that is associated with described a plurality of networks of different type business, be described a plurality of networks of different type traffic assignments resources.
2. server according to claim 1 is characterized in that described network connector comprises ethernet connector.
3. server according to claim 1 is characterized in that, described a plurality of networks of different type business comprise at least two in public Ethernet service, unloading business, storage service and the long-range direct memory access (DMA) business.
4. server according to claim 1 is characterized in that described processor comprises single integrated chip.
5. server according to claim 1 is characterized in that, described processor comprises 2 layer network interface cards, transmission control protocol processor and upper-layer protocol processor.
6. server according to claim 5 is characterized in that, described transmission control protocol processor provides 3 layers of processing and 4 layers of processing.
7. server according to claim 5 is characterized in that, by TCP unloading is professional, internet small computer system interface is professional and RDMA at least two in professional share described TCP processors.
8. server according to claim 5 is characterized in that, described upper-layer protocol processor provides internet small computer system interface to handle.
9. server according to claim 5 is characterized in that, described upper-layer protocol processor provides RDMA to handle.
10. server according to claim 1 is characterized in that, further comprises:
Be connected to the server admin agency of described processor.
11. server according to claim 1 is characterized in that, described server admin agency is connected to keyboard and/or video and/or mouse service.
12. server according to claim 1 is characterized in that, further comprises:
Be connected to a plurality of services of described unified driver.
13. server according to claim 12 is characterized in that, described a plurality of services comprise at least two in socket service, the service of SCSI mini-port, RDMA service and keyboard and/or video and/or the mouse service.
14. server according to claim 1 is characterized in that, described unified driver is connected to software TCP processor and socket service switch,
Wherein said software TCP processor is connected to described socket service switch, and
Wherein said socket service switch is connected to socket service.
15. server according to claim 1 is characterized in that, described processor or PCI bridge are determined the specific service which access in the networks of different type business is provided by described server.
16. server according to claim 15 is characterized in that, described specific service comprises at least one in socket service, the service of SCSI mini-port, RDMA service and keyboard and/or video and/or the mouse service.
17. server according to claim 1 is characterized in that, described processor, described PCI bridge or described unified driver provide uniform data and control path.
18. a method for connecting network is characterized in that, comprising:
(a) handle a plurality of networks of different type business via 2 layers of connector;
(b) the networks of different type business in the processing single-chip;
(c) determine that in the described networks of different type business which is via forms data path access software service.
19. method according to claim 18 is characterized in that, described a plurality of networks of different type business comprise at least two in public Ethernet service, unloading business, storage service, interprocess communication business and the management service.
20. method according to claim 18 is characterized in that, described 2 layers of connector are single 2 layers of connectors.
21. method according to claim 18 is characterized in that, described step (c) comprises and adopts time division multiplex to determine that in the described networks of different type business which is via the described software service of described forms data path access.
22. method according to claim 18 is characterized in that, described step (c) comprises the fixed resource between the described networks of different type business of dynamic assignment.
23. method according to claim 18 is characterized in that, further comprises:
(a) provide the driver that is associated with described a plurality of networks of different type business via consolidator.
24. a method of network connection is characterized in that, comprising:
(a) handle a plurality of networks of different type business via the individual ethernet connector;
(b) use 2 layers of processor, 3 layers of processor, 4 layers of processor and the described a plurality of networks of different type business of upper-layer protocol processor processing; And
(c) provide uniform data and control path.
25. method according to claim 24 is characterized in that, described 2 layers of processor comprise single 2 layer network interface cards.
26. method according to claim 24 is characterized in that, described 3 layers of processor and 4 layers of processor are combined into single TCP processor.
27. method according to claim 24 is characterized in that, described upper-layer protocol processor comprises at least one in internet small computer system interface processor and the long-range direct memory access (DMA) processor.
28. method according to claim 24 is characterized in that, further comprises:
(a) provide the driver that is associated with described a plurality of networks of different type business via single unified driver.
CNB038201275A 2002-08-29 2003-08-29 System and method for network interfacing in a multiple network environment Expired - Fee Related CN100472450C (en)

Applications Claiming Priority (25)

Application Number Priority Date Filing Date Title
US10/230,643 2002-08-29
US10/230,643 US7295555B2 (en) 2002-03-08 2002-08-29 System and method for identifying upper layer protocol message boundaries
US60/407,165 2002-08-30
US60/408,207 2002-09-04
US60/408,617 2002-09-06
US60/410,022 2002-09-11
US60/411,294 2002-09-17
US60/419,354 2002-10-18
US60/420,901 2002-10-24
US60/425,959 2002-11-12
US10/298,817 2002-11-18
US10/302,474 2002-11-21
US60/434,503 2002-12-18
US60/437,887 2003-01-02
US10/336,983 2003-01-06
US10/337,029 2003-01-06
US60/439,951 2003-01-14
US60/442,360 2003-01-24
US60/456,265 2003-03-20
US60/456,260 2003-03-20
US60/456,266 2003-03-20
US60/456,322 2003-03-20
US60/477,279 2003-06-10
US60/478,106 2003-06-11
US10/652,327 2003-08-29

Publications (2)

Publication Number Publication Date
CN1757016A CN1757016A (en) 2006-04-05
CN100472450C true CN100472450C (en) 2009-03-25

Family

ID=36689400

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB038201275A Expired - Fee Related CN100472450C (en) 2002-08-29 2003-08-29 System and method for network interfacing in a multiple network environment

Country Status (1)

Country Link
CN (1) CN100472450C (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE112011106016T5 (en) * 2011-12-23 2014-09-11 Intel Corporation Common send queue

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6421728B1 (en) * 1997-12-31 2002-07-16 Intel Corporation Architecture for communicating with and controlling separate upstream and downstream devices

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6421728B1 (en) * 1997-12-31 2002-07-16 Intel Corporation Architecture for communicating with and controlling separate upstream and downstream devices

Also Published As

Publication number Publication date
CN1757016A (en) 2006-04-05

Similar Documents

Publication Publication Date Title
US9088451B2 (en) System and method for network interfacing in a multiple network environment
US8046465B2 (en) Flow scheduling for network application apparatus
US7934021B2 (en) System and method for network interfacing
US6389432B1 (en) Intelligent virtual volume access
US20030099254A1 (en) Systems and methods for interfacing asynchronous and non-asynchronous data media
US20030097481A1 (en) Method and system for performing packet integrity operations using a data movement engine
CN102185833B (en) Fiber channel (FC) input/output (I/O) parallel processing method based on field programmable gate array (FPGA)
US20090147677A1 (en) System, method, and apparatus for load-balancing to a plurality of ports
US20040260841A1 (en) Method, apparatus, and system for internet protocol communication over intelligent platform management bus
EP1570369A2 (en) An apparatus and method for receive transport protocol termination
US20110314171A1 (en) System and method for providing pooling or dynamic allocation of connection context data
US6742075B1 (en) Arrangement for instigating work in a channel adapter based on received address information and stored context information
EP1460806A2 (en) System and method for network interfacing in a multiple network environment
EP1540473B1 (en) System and method for network interfacing in a multiple network environment
JP2000235536A (en) Data communication system and device
CN100472450C (en) System and method for network interfacing in a multiple network environment
EP1158750B1 (en) Systems and method for peer-level communications with a network interface card
JPS59176952A (en) Communication control system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20090325

Termination date: 20170829

CF01 Termination of patent right due to non-payment of annual fee