CN108604992A - The system and method switched using the software definition between the lightweight virtual machine of host kernel resources - Google Patents

The system and method switched using the software definition between the lightweight virtual machine of host kernel resources Download PDF

Info

Publication number
CN108604992A
CN108604992A CN201780009111.0A CN201780009111A CN108604992A CN 108604992 A CN108604992 A CN 108604992A CN 201780009111 A CN201780009111 A CN 201780009111A CN 108604992 A CN108604992 A CN 108604992A
Authority
CN
China
Prior art keywords
message
data packet
node
nodes
destination
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.)
Granted
Application number
CN201780009111.0A
Other languages
Chinese (zh)
Other versions
CN108604992B (en
Inventor
拉各韦卓·克沙瓦穆西
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.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
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
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Publication of CN108604992A publication Critical patent/CN108604992A/en
Application granted granted Critical
Publication of CN108604992B publication Critical patent/CN108604992B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/02Details
    • H04L12/04Switchboards
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • G06F2009/45595Network integration; Enabling network access in virtual machine instances

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The present invention provides a kind of application layer crossbar switches for allowing to be communicated across container/lightweight virtual machine.The application layer crossbar switch or software cross switch or software definition switch provide unified communication interface for application layer, have been abstracted connection management and details that message sends and receives.Software definition switch/application layer crossbar switch can be used for connecting cyberspace using domain socket and, to the isolation this point of message exchange in host, improve the message exchange performance of virtual machine in host without influencing.

Description

It is using what the software definition between the lightweight virtual machine of host kernel resources switched System and method
Technical field
Theme described herein relates generally to communication data processing, more specifically, for improving in lightweight virtual machine Network capabilities performance.
Background technology
In computer realm, virtual machine (virtual machine, VM) is the emulation to particular computer system.Virtually Machine is operated based on true or hypothetical computer computer architecture and function, and realization may relate to specialized hardware, software Or combination.It is well known that operation virtual machine has many benefits.Virtual machine makes better use of hardware, be easy to backup and It exchanges, and business is isolated from each other.But operation virtual machine also has disadvantage.Virtual machine image is very clumsy.In addition, prior It is that virtual machine needs considerable resource, because they simulate hardware and have run full stack operation system.
By using linux container, a kind of lightweight substitute is provided to substitute ripe virtual machine, while can retain maturation The advantages of virtual machine.Container is OS-Level virtual environment, multiple independent for being run on single Linux hosts Linux system, otherwise referred to as lightweight virtualize or lightweight virtual machine.RunC, Docker and Warden are lightweight void Some examples of quasi- machine (container), can be used for building " platform services " (Platform as a Solution, PaaS). RunC/Docker/Warden is the scheme based on linux kernel NameSpace and CGroup, by wieldy console/ API/ picture formats are abstracted complicated core A PI, and provide and be abstracted on demand to calculate storage and network capabilities.
Container networking can probably be considered as creates consistent network environment to one group of container.Stacking network can be used to realize for this, There are numerous embodiments, such as Docker default networks pattern, weave, flannel and socketplane for stacking network.Institute The main advantage of these stacking networks is without changing application code, using can dispose as it is.Network namespace connects Mode to physical network device is also another part of container networking.
There are multiple permission network namespaces to be communicated with network hardwares such as veth, OpenVSwith, domain sockets Linux kernel module.NameSpace is a feature of linux kernel, allows to separate one group of process, makes these processes It " cannot see " resource in other groups.According to the linux kernel document of network namespace, Network Isolation rank is not being damaged In the case of, network namespace can be communicated by veth withs with and/or domain socket.Fig. 1 (a) is shown using veth pairs The network namespace communicated.Veth is to being a kind of virtual unit for the similar Ethernet that can be used in container.veth To capturing ethernet frame, the frame of capture can be sent to destination by bridge or router.Domain socket is a kind of efficient light Magnitude interprocess communication (inter process communication, IPC).Fig. 1 (b) is shown to be carried out using domain socket The network namespace of communication.Domain socket can be controlled by file permission, so the TCP that can be connected than anyone Port is safer, therefore TCP port needs further security protection.
Fig. 2 shows the network settings in Docker.Fig. 2 shows virtual machine/containers to be used for solving the logical of network problem Use method.The universal method is to capture Ethernet data bag from virtual units such as veth, then by bridge/router by data Packet is sent in the required container in same host or different hosts.Veth to be it is a kind of can be used in container it is similar with The virtual unit netted very much.Fig. 3 shows that the inside in Docker is arranged.As shown in figure 3, each generate/instantiation container All using veth to being connected to Linux bridges.Veth pairs of container end is known as eth0, and the veth of Linux bridges side is to being known as Vethxx, such as vethab or vethbb.Similar veth pairs is connected between host and Linux bridges.
As generally known, bridge operates in layer 2 (Layer 2, L2) and independently of agreement.L2 network layers are responsible for Physical addressing, error correction and be medium ready message.Bridge is to be connected to two independent network segments in a manner of independently of agreement A kind of mode together.Data packet is based on ethernet address rather than IP address (such as router) forwards.Forwarding is in layer 2 (Layer2, L2) is completed, therefore all agreements can pass through bridge transparent transmission.All-network flow all passes through Linux bridges Or other bridges of configuration, such as OpenVSwitch (OVS).Using IP tables (as shown in Figure 3), so as to by each container end Mouth is mapped to host port.Each generate/instantiation IP address can change when each container is restarted, this also brings along Problem, because the business in a container must obtain the new IP address for accessing these business.
Fig. 4 shows the communication in stacking network or L2/L3 schemes.Stacking network is built upon the meter on another network Calculation machine network.Node in stacking network can be considered as through virtual or logical link connection, wherein each link corresponds to bottom A path in layer network, the path may pass through many physical links.Scheme based on L2 means from similar virtual ether The equipment of net captures L2 frames, then sends these frames by other transmission mechanisms such as TCP/UDP/SecureTCP.
Fig. 4 is shown captures data packet as available in the prior art in Ethernet grade (virtual ethernet equipment), then By transmission such as TCP/UDP come the method for transmission data packet.This is known as stacking network or L2/L3 schemes.However, shown in Fig. 4 Method there is a problem of related to network efficiency and Operating Complexity.Network efficiency is very low because data L2 (network stack 2) layer is captured, then Reseal to send it to correct destination or be route by bridge.In multi-host network L2 schemes can cause the operational issues such as configuration error, need digerait's commissioning.
Fig. 5 shows the details across the message exchange of two hosts.As shown in Fig. 5 (a):Sender applies (application, App) creates a socket for arriving destination, and sender's application formatted data is simultaneously connect using socket Mouth sends message, and the TCP/IP stacks of kernel are further processed and are sent to network interface card.The sides App are being received, network interface card receives data, TCP/IP stacks are forwarded to, TCP/IP stacks are sent to App again.It is used as using superposition when by the default network pattern of lightweight virtual machine When the traditional mode of network, the step needs performed by above-mentioned Fig. 5 (a) are further processed, to need additional CPU and storage Device as shown in Fig. 5 (b), and then influences network performance.
Therefore, in view of the above, it is obvious that when using lightweight virtual machine, calculate and storage capacity Performance is unaffected, but the performance of network capabilities is severely limited.
The drawbacks described above for the lightweight virtual machine realized in terminal device now is just for the sake of general introduction legacy system/machine A part of problem of system/technology is non-detailed.Other problems of legacy system/mechanism/technology and described herein various The corresponding benefit of non-limiting embodiment becomes apparent from when reading is described below.
Invention content
The content of present invention is for introducing concept related with the performance for improving network capabilities in lightweight virtual machine, below Detailed description in further describe these concepts.The content of present invention is not intended to confirm the basic of theme claimed Feature is also not intended to assist in or limit the range of theme claimed.
Improve the performance of network capabilities the main purpose of the present invention is to provide a kind of mechanism, in particular container/light The performance of magnitude virtual machine, to solve above-mentioned technical problem.
Therefore, the present invention provides a kind of application layer crossbar switch for allowing to be communicated across container/lightweight virtual machine.Institute It states application layer crossbar switch or software cross switch or software definition switch provides unified communication interface for application layer, be abstracted The details that connection management and message send and receive.Software definition switch/application layer crossbar switch can be with using domain socket For connecting cyberspace, to the isolation this point of message exchange in host, the message of virtual machine in host is improved without influencing Switching performance.
In one embodiment, software cross switch/application layer crossbar switch/software definition switch is to be using biography Send the software switch of data.Using being abstract from connection management.
In one embodiment, the present invention provides the node in a kind of communication system.The node include processor and Memory, the memory are coupled to the processor, the processor for execute be present in it is multiple in the memory Module.The multiple module includes at least one interface module and at least one processing module.The interface module is for sending And/or at least one message/data packet is received, interface is by residing in the application initialization of at least one of described node.It is described Processing module is used to provide at least one connection management to the other nodes of one or more of the communication system to realize system One communication, the connection management are obtained based at least one of following:IP address, shared memory key, public correspondence port, And their arbitrary combination.
In one embodiment, the present invention provides the node in a kind of communication system.The node include processor and Memory, the memory are coupled to the processor, the processor for execute be present in it is multiple in the memory Module.The multiple module includes at least one crossbar switch processing module.The crossbar switch processing module is used for:Using extremely A few crossbar switch bank interface receive by reside in the application initialization of at least one of described node at least one message/ Data packet;The destination of the message/data packet is verified, the destination is same node, different nodes or its arbitrary combination; If the destination of the message/data packet be same node, in the same node establishment/use open field set Connect word connection;Or the IP address based on the one or more of the other node received in the message/data packet disappears described Breath/data packet is transmitted to other nodes, and other nodes include lightweight virtual machine process.
In one embodiment, the present invention provides a kind of communication system.The communication system includes:Multiple main frames equipment and Multiple server apparatus;Processor;And at least one crossbar switch of the embedded processor.The crossbar switch with it is described Host equipment and server apparatus interconnection, for the communication between the host equipment and/or the server apparatus Unified communication interface is provided.The crossbar switch is used for:It is received using at least one crossbar switch bank interface by residing in State at least one message/data packet of at least one of host application initialization;Verify the purpose of the message/data packet Ground, the destination are the host, the server or its arbitrary combination;If the purpose of the message/data packet Ground is same node, the then establishment/connected using open field socket in same host equipment and/or server apparatus;Or Based on the IP address of at least one other host/server equipment received in the message/data packet, by the message/ Data packet is transmitted to other host/server equipment, other host/server equipment include lightweight virtual machine into Journey.
In one embodiment, the present invention provides a kind of method that the node by communication system executes.The method packet It includes:At least one interface is sent and/or is received by residing in at least one of at least one of described node application initialization Message/data packet;At least one processing module provides at least one to the other nodes of one or more of the communication system Connection management is to realize that Unified Communication, the connection management are obtained based at least one of following:IP address, shared memory are close Key, public correspondence port and their arbitrary combination.
In one embodiment, the present invention provides a kind of method that the node by communication system executes.The method packet It includes:It is received using at least one crossbar switch bank interface by residing at least one of described node application initialization at least One message/data packet;Verify the destination of the message/data packet, the destination be same node, different nodes or its Arbitrary combination;If the destination of the message/data packet is same node, creates and use in the same node Open field socket connects;Or the IP address based on the one or more of the other node received in the message/data packet The message/data packet is transmitted to other nodes, other nodes include lightweight virtual machine process.
Compared with prior art, it is a major advantage of the invention that a kind of software cross switch of present invention offer or application Crossbar switch or software definition switch provide network abstraction to the application run in lightweight virtual machine, simplify application Exploitation/dispose and improve network performance.
In addition, the present invention is suitable for physical machine (physical machine, PM), can apply to run multiple virtual machines Physical machine (physical machine running many virtual machine, PM-VM), can apply to run it is multiple Virtual machine and each virtual machine run physical machine (the Physical machine running many of multiple lightweight virtual machines virtual machine,each virtual machine running many Light weight virtual Machines, PM-VM-LVM), moreover it is possible to apply to PM-LVM deployment combinations.
Compared with prior art, the network performance that the present invention realizes is more preferable, because of lightweight virtual machine communication in host Expense is seldom.
Compared with prior art, through the invention, multi-host network is simplified, because local crossbar switch is socketed using domain Word forwards the messages to gateway crossbar switch, is encapsulated without additional message.
Compared with prior art, through the invention, the networks configuration of territories such as IPTable rules, port mapping are not necessarily to, by subtracting Few operation overhead remains to realize better network performance.
In addition, compared with prior art, through the invention, in host across the message exchange of lightweight virtual machine and Speech, the performance provided is identical as the performance that local domain socket provides, but due to operation and maintenance demand is such as logged in, it can It can decline 10%.In addition, for the message exchange across host lightweight virtual machine, performance will be better than increasing income in the market Stacking network scheme.
Above-mentioned various options related with first embodiment and preferred embodiment are also applied for other embodiment party Formula.
Description of the drawings
The detailed description is given with reference to the accompanying drawings.In the accompanying drawings, it is referred to described in the leftmost digital representation of Ref. No. Number first appears in the figure.All attached drawings refer to the component of same characteristic features using same numbers.
Fig. 1 shows that the network namespace for using veth couples of (a) and the domain (b) socket communicates;
Fig. 2 shows the network settings in Docker;
Fig. 3 shows that the inside in Docker is arranged;
Fig. 4 shows stacking network or L2/L3 schemes;
Fig. 5 shows the message exchange across two host (a) sender sides and recipient, and (b) message sink Processing cost;
Fig. 6 show carried out using application layer crossbar switch according to the embodiment of this theme it is empty across container/lightweight Quasi- machine communication;
Fig. 7 shows the processing carried out according to the crossbar switch design and use crossbar switch of the embodiment of this theme;
Fig. 8 shows the operation order flow in the same host of the embodiment according to this theme, different vessels scene;
Fig. 9 shows the operation order flow in the different hosts of the embodiment according to this theme, different vessels;
Figure 10 shows the node in a kind of communication system according to the embodiment of this theme;
Figure 11 shows the node in a kind of communication system according to the embodiment of this theme;
Figure 12 shows the method that a kind of node by communication system according to the embodiment of this theme executes.
It should be understood that attached drawing is in order to illustrate idea of the invention, may be not drawn on scale.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention is clearly retouched It states.Obviously, described embodiment is only a part of the embodiment of the present invention, instead of all the embodiments.Based on of the invention real Example is applied, all other embodiment obtained by those of ordinary skill in the art without making creative efforts all belongs to In the scope of protection of the invention.
The present invention can be implemented in several ways, can be embodied as process, device, system, synthetic, computer-readable The computer-readable medium of storage medium etc., or in which program instruction pass through the computer network that light or electronic communication link are sent Network.In the present specification, any other form that these embodiments or the present invention can take all can be described as technology.In general, The sequence of the step of during disclosed in can changing within the scope of the invention.
The detailed description of one or more embodiments of the invention is provided below and illustrates the attached drawing of the principle of the invention.This Invention is described in conjunction with these embodiments, but the present invention is not limited to any embodiments.The scope of the present invention is only wanted by right Limitation is asked, and the present invention includes many substitutes, modification and equivalent.Numerous specific details are set forth in following description, To understand thoroughly the present invention.These detailed contents are used for example, in some or all of these no details In the case of, the present invention can still be implemented according to claims.For clarity, technical field for the present invention is not described in detail In known technical data, unnecessary obscure in order to avoid being generated to the present invention.
In detailed description below, numerous specific details are set forth to understand thoroughly the present invention.However, art technology Personnel will be understood that also the present invention may be practiced without these details.In other cases, be not described in detail well known method, Process, component, module, unit and/or circuit, in order to avoid obscure the present invention.
Although the embodiment of the present invention is unrestricted in this regard, such as " processing ", " operation ", " calculatings ", " true are used Calmly ", the discussion of the terms such as " foundation ", " analysis ", " inspection " can refer to computer, computing platform, computing system or other electrometers The operation and/or processing for calculating equipment, will be expressed as the data of physical quantity (such as electricity) in computer register and/or memory Handle and/or be converted to computer register and/or memory or can store instruction to execute the other non-of operation and/or processing The other data for being equally expressed as physical quantity in instantaneity information storage medium.
Although the embodiment of the present invention is unrestricted in this regard, terminology used in this article " many " and " multiple " It may include such as " several " or " two or more ".Term " many " or " multiple " can be used to describe two throughout the specification A or more component, equipment, component, unit, parameter etc..Unless explicitly stated otherwise, otherwise embodiment of the method as described herein is not It is limited to particular order or order.In addition, described a part of embodiment of the method or its element can synchronize while or parallel hair Raw or execution.
When using lightweight virtual machine, calculates and the performance of storage capacity is unaffected, but the performance of network capabilities is tight It is limited again.
Therefore, the present invention provides a kind of application layer crossbar switches for allowing to be communicated across container/lightweight virtual machine. Application layer crossbar switch or software cross switch or software definition switch as shown in FIG. 6 provide unified communication for application layer and connect Mouthful, it has been abstracted connection management and details that message sends and receives.Software definition switch/application layer crossbar switch is covered using domain It connects word and can be used for connecting cyberspace without influencing the isolation this point to message exchange in host, improve virtual in host The message exchange performance of machine.
In one embodiment, software cross switch/application layer crossbar switch/software definition switch is to be using biography Send the software switch of data.Using being abstract from connection management.
It discloses for improving the software definition switch in virtual machine between the lightweight virtual machine of the performance of message exchange System, method and node.
Although describing system and the side of the software definition switch between the lightweight virtual machine using host kernel resources The various aspects of method, but the present invention can realize in any amount of different computing systems, environment and/or configuration, with shown below Example sexual system, equipment/node/device and method context in describe embodiment.
Hereafter the embodiment of the present invention is explained by means of graphical representation of exemplary and one or more examples.However, these examples Property diagram and example be for illustration purposes, to more fully understand the present invention, to be not necessarily to be construed as the limit to the scope of the invention System.
It is carried out according to the crossbar switch design and use crossbar switch of the embodiment of this theme referring now to Figure 7, showing Processing.
In one embodiment, as shown in fig. 7, crossbar switch may be designed as two parts:Crossbar switch library and crossbar switch Process.
In one embodiment, crossbar switch library provides a kind of interface, the i.e. application program for sending and receiving message Interface (application program interface, API).Using that must initialize this library, exposure initializes API It is for this purpose.Crossbar switch library will use shared memory in inside, so as to crossbar switch library and crossbar switch into Message is quickly exchanged between journey.
In one embodiment, crossbar switch process provides the practical connection management between other virtual machines.It is each to intersect Switch process must configure relevant details, such as IP address, shared memory key and common user communication port.It is general and Speech, common user communication port is set as 9999.
In one embodiment, each lightweight virtual machine can dispose at least one crossbar switch, and referred to as local intersection is opened It closes.All applications to be run in the lightweight virtual machine must all be linked with crossbar switch library.
All applications to be run in the lightweight virtual machine are all linked with crossbar switch library.For example, g++-g Myapp.cpp-L/home/abc/lib-lCrossbar-o myApp, wherein myApp (application) are linked to crossbar switch library simultaneously Using its API come send/receive message.
In one embodiment, at least one crossbar switch, referred to as gateway crossbar switch can be disposed in host-level.Each originally Ground crossbar switch is all configured with the address of gateway crossbar switch.Those skilled in the art can pay attention to and understand, Mei Geqi Dynamic lightweight virtual machine must have unique IP in network settings.
In one embodiment, if message is assigned to the process in different lightweight virtual machines, crossbar switch Library can have the information (destination address) for placing the message in local crossbar switch process queue.In one example, if Target IP is different from current hosts IP, then destination is not host.Therefore, it is necessary to place the message in network queue and non-local friendship Fork closes process queue.
In one embodiment, crossbar switch process is by polling message, when there is message event, can check that purpose process is In same host still in different hosts.
In one embodiment, if destination is same host, crossbar switch process is light by establishment/use and purpose The socket of open field connection between magnitude virtual machine process.
In one embodiment, if destination is different hosts, crossbar switch process can be forwarded the message at this The gateway crossbar switch process run in host.
In one embodiment, gateway crossbar switch is by the routing configured with other host gateways.Gateway crossbar switch will Other host gateways are found based on destination IP, and message is forwarded using TCP transmission.Gateway crossbar switch will use network Host mode, to obtain better performance.
In one embodiment, in case of any update, new gateway crossbar switch is such as added, then the information will be automatic It is synchronized to entire gateway crossbar switch.It may be used for carrying out this synchronization there are one time window, during this period, using possible It will receive the failed message not arrived at.It adds new gateway crossbar switch and is considered as network management activities, therefore need Application is designed by certain notice business, to reach same purpose.
Referring again to FIGS. 7, in one embodiment, crossbar switch process can have shared memory managing process, domain to cover Connect word connection Thread process, domain socket watcher thread process, TCP socket connection Thread process, TCP socket watcher thread Process transmits and receives Thread process.
In one embodiment, shared memory managing process can create/be attached to given shared memory, from altogether It enjoys memory area and creates queue, and will apply and be mapped to queue.
In one embodiment, domain socket connection Thread process can be established with other local lightweight virtual machine processes Connection, and if purpose is applied in the lightweight virtual machine of different hosts, socket connection Thread process in domain can will answer It is forwarded to gateway crossbar switch with message.
In one embodiment, domain socket watcher thread process can receive new connection request and handle the new connection request For subsequently using.
In one embodiment, TCP socket connection Thread process can be with other gateway crossbar switches in other hosts Connection is established, and updates routing (mapping of gateway IP, IP mask).
In one embodiment, TCP socket watcher thread process can receive new connection request and handle the new connection and asks It asks, such as prepares data structure, for subsequently using.
In one embodiment, the browsable data structure by monitoring/connection thread creation of Thread process, inspection are transmitted and received The suitable socket handle of rope and transmission/reception message.
Referring again to FIGS. 7, in one embodiment, crossbar switch library can have shared memory managing process and interface.
In one embodiment, shared memory managing process can create/be attached to given shared memory, from altogether It enjoys memory area and creates/be attached queue, and App is mapped to queue.
In one embodiment, interface is for sending or receiving or initiate or abandoning received data packet or message.
Referring now to Figure 8, showing the operation in same host, different vessels scene of the embodiment according to this theme Sequence flow.
Referring now to Figure 9, showing the operation order in different hosts, different vessels of the embodiment according to this theme Flow.
Referring now to Figure 10 and Figure 11, the node in a kind of communication system according to the embodiment of this theme is shown 1000.In one embodiment, node 1000 is disclosed.Although illustrating this master by realizing the present invention in node 1000 Topic, but it is understood that, the present invention can also realize in various computing systems, such as laptop computer, desk-top meter Calculation machine, laptop, work station, mainframe computer, server, network server etc..It will be understood that node 1000 can be by Multiple users reside in application access in Database Systems.The example of node 1000 can include but is not limited to portable meter Calculation machine, personal digital assistant, hand-held node, sensor, router, gateway and work station.Node 1000 can communicatedly coupling Other nodes or a certain node or device are closed to form network (not shown).The example of other nodes or a certain node or device It can include but is not limited to portable computer, personal digital assistant, hand-held node, sensor, router, gateway and work It stands.
In one embodiment, network (not shown) can be wireless network, cable network or combinations thereof.The network can be real It is now one of different types of network, such as GSM, CDMA, LTE, UMTS, Intranet, LAN (local area Network, LAN), wide area network (wide area network, WAN), internet etc..The network can be dedicated network or shared Network.Shared network representation uses such as hypertext transfer protocol (Hypertext Transfer Protocol, HTTP), biography Transport control protocol view/mutual procotol (Transmission Control Protocol/Internet Protocol, TCP/ IP), the various agreements such as Wireless Application Protocol (Wireless Application Protocol, WAP) are in communication with each other not The combination of same type network.In addition, the network may include various network nodes, including router, bridge, server, calculating section Point, memory node etc..
Node 1000 may include processor 1002, interface 1004 and memory 1006.Processor 1002 can realize to be one A or multi-microprocessor, microcomputer, microcontroller, digital signal processor, central processing unit, state machine, logic circuit And/or any node based on operational order manipulation signal.At least one processor is stored in memory for obtaining and executing Computer-readable instruction or module in 1006, etc..
Interface (I/O interfaces) 1004 may include various software and hardware interfaces, such as network interface, graphical user interface Deng.I/O interfaces allow Database Systems, first node, second node and third node directly to be interacted with user.In addition, I/O Interface can enable node 1000 and other nodes, node or the calculating such as network server and Outside data services device (not shown) Node is communicated.I/O interfaces can promote a variety of communications in multiple network and protocol type, including GSM, CDMA, LAN, The wireless networks such as the cable networks such as cable and WLAN, honeycomb or satellite.I/O interfaces may include one or more ports, be used for Multiple nodes are connected to each other or are connected to another server.I/O interfaces can provide user by being supplied to the screen of interface With the interaction between node 1000.
Memory 1006 may include that any computer-readable medium known in the art, computer-readable medium include for example Volatile memory, such as static RAM (static random access memory, SRAM) and dynamic random Access memory (dynamic random access memory, DRAM) and/or nonvolatile memory, such as read-only storage Device (read only memory, ROM), erasable programmable ROM, flash memory, hard disk, CD and tape.Memory 1006 can wrap Multiple instruction, module or application are included, to perform various functions.Memory includes executing particular task or realization specific abstract data Routine, program, object, component, data structure of type etc..
In one embodiment, the present invention provides the node 1000 in a kind of communication system.Node 1000 includes processor 1002 and memory 1006, memory 1006 is coupled to processor 1002, and processor 1002 is present in memory for executing Multiple modules in 1006.Multiple modules include at least one interface module 1008 and at least one processing module 1010.Interface Module 1006 is applied for sending and/or receiving at least one message/data packet, interface by residing at least one of node Initialization.Processing module 1010 be used for the other nodes of one or more of communication system provide at least one connection management with Realize Unified Communication;Connection management is obtained based at least one of following:IP address, shared memory key, public correspondence end Mouth and their arbitrary combination.
In one embodiment, the present invention provides the node 1000 in a kind of communication system.Node 1000 includes processor 1002 and memory 1006, memory 1006 is coupled to processor 1002, and processor 1002 is present in memory for executing Multiple modules in 1006.Multiple modules include at least one crossbar switch processing module 1102.Crossbar switch processing module 1102 using at least one crossbar switch bank interface for being received by residing at least one of node application initialization extremely A few message/data packet;Verify the destination of message/data packet, destination is same node, different node or its arbitrary group It closes;If the destination of message/data packet is same node, the establishment/connected using open field socket in the same node It connects;Or message/data packet is transmitted to by the IP address based on the one or more of the other node received in message/data packet Other nodes, other nodes include lightweight virtual machine process.
In one embodiment, the present invention provides a kind of communication system.Communication system includes:Multiple main frames equipment and multiple Server apparatus;Processor;And at least one crossbar switch of embedded processor.Crossbar switch and host equipment and server Apparatus interconnection provides unified communication interface for the communication between host equipment and/or server apparatus.Crossbar switch is used In:It is received using at least one crossbar switch bank interface by residing in at least one of at least one of host application initialization Message/data packet;The destination of message/data packet is verified, destination is host, server or its arbitrary combination;If message/ The destination of data packet is same node, the then establishment/be socketed using open field in same host equipment and/or server apparatus Word connects;Or the IP address based on at least one other host/server equipment received in message/data packet, it will disappear Breath/data packet is transmitted to other host/server equipment, and other host/server equipment include lightweight virtual machine process.
In one embodiment, interface module 1008 is additionally operable to be discharged at least with using associated message/data packet In at least one queue in one shared memory.
In one embodiment, processing module 1010 is additionally operable to:Extract queue in apply associated message/data It wraps, and verifies the destination of message/data packet, destination is same node, different nodes or its arbitrary combination.
In one embodiment, if the destination of message/data packet is same node, processing module is additionally operable at this Establishment in same node/connected using open field socket.
In one embodiment, if the destination of message/data packet is different nodes, processing module is additionally operable to be based on The IP address of destination transmits message/data packet.
In one embodiment, IP address-based message/data packet is using predefine/being pre-configured and be pre-stored in processing At least one of module route to transmit, which follows TCP transfer mechanisms.
In one embodiment, processing module 1010 is used for the case where not influencing to the isolation of message exchange in host Under, connect cyberspace using domain socket.
In one embodiment, IP address associated with other nodes, shared memory key, public correspondence port or Their arbitrary combination is pre-stored/is pre-configured to be in the memory of node.
In one embodiment, the message/data packet of reception is stored at least one at least one shared memory In queue.
In one embodiment, crossbar switch processing module 1102 is additionally operable to extraction storage in the queue associated and disappears Breath/data packet, to verify the destination of message/data packet.
In one embodiment, using predefine/be pre-configured and be pre-stored in crossbar switch processing module at least one Message/data packet is transmitted to other nodes by a routing.The routing follows TCP transfer mechanisms.
In one embodiment, crossbar switch processing module 1102 is additionally operable to establish with other nodes in communication system and connect It connects, other nodes preferably include lightweight virtual machine process, and connection is obtained based at least one of following:IP address is shared Memory key, public correspondence port and their arbitrary combination.
In one embodiment, other nodes in communication system are used for after receiving message/data packet:By being based on The message/data packet of reception creates at least one data structure to handle new connection request;Update is predefined/it is pre-configured and pre- The routing being stored in crossbar switch processing module, it is preferable that message/data packet of the routing based on reception passes through gateway IP, IP Mask or its mapping arbitrarily combined are updated;The created data structure of scanning is to extract suitable socket handle;From And if the destination of message/data packet is same node, the establishment/connected using open field socket in the same node It connects;Or message/data packet is transmitted to by the IP address based on the one or more of the other node received in message/data packet Other nodes.
Referring now to Figure 12, showing the side that a kind of node by communication system according to the embodiment of this theme executes Method.This method can be described in computer executable instructions this general contexts.In general, computer executable instructions can wrap Include execute specific function or realize the routine of particular abstract data type, program, object, component, data structure, flow, module, Function etc..This method can also be implemented in a distributed computing environment, and in a distributed computing environment, function is by passing through communication network The remote processing devices of link execute.In a distributed computing environment, computer executable instructions can be located locally and remotely count Calculation machine storage medium both, including memory storage device.
The sequence of description method should not be construed as limiting, and can combine any amount of the method step in any order To realize this method or alternative.In addition, can be in the case where not departing from the protection domain of theme described herein from this method Middle deletion single step.In addition, this method can be implemented in any suitable hardware, software, firmware or combinations thereof.However, For convenience of description, in the embodiment being described below, it is believed that this method is implemented in above-mentioned node 1000.
Referring now to Figure 12, disclosing a kind of method that the node 1000 by communication system executes.
At step 1202, at least one by residing in the application initialization of at least one of node 1000 is sent or received A message/data packet.Message/data packet is sent or is received by least one interface of node 1000.
At step 1204, at least one of node 1000 will be come/is stored in using associated message/data packet In at least one queue in shared memory.
At step 1206, extraction storage in the queue with apply associated message/data packet.
At step 1208, node 1000 verify queue in application associated message/data packet destination.Mesh Ground be same node, different node or its arbitrary combine.
In one embodiment, if the destination of message/data packet is same node, processing module is additionally operable at this Establishment in same node/connected using open field socket.
In one embodiment, if the destination of message/data packet is different nodes, processing module is additionally operable to be based on The IP address of destination transmits message/data packet.
At step 1210, at least one connection is established to realize system with the other nodes of one or more of communication system One communication;Connection management is obtained based at least one of following:IP address, shared memory key, public correspondence port, and Their arbitrary combination.In one embodiment, and in the case of influence the isolation of interacting message in host, using one or Multiple domain sockets connect cyberspace.IP address associated with other nodes, shared memory key, public correspondence end Mouth or their arbitrary combination can be pre-stored/be pre-configured to be in the memory of node.
In one embodiment, other nodes preferably include lightweight virtual machine process, and connection is based on following at least one And obtain:IP address, shared memory key, public correspondence port and their arbitrary combination.Its in communication system Its node creates at least one data structure to locate after receiving message/data packet, by the message/data packet based on reception Manage new connection request;Update the routing for predefine/being pre-configured and being pre-stored in crossbar switch processing module, it is preferable that road It is updated by gateway IP, IP mask or its mapping arbitrarily combined by the message/data packet based on reception;Scanning is created Data structure to retrieve suitable socket handle;If the destination of message/data packet is same node, same at this Establishment in node/connected using open field socket;Or based on the one or more of the other section received in message/data packet Message/data packet is transmitted to other nodes by the IP address of point.
At step 1212, the IP address based on destination transmits message/data packet.In one embodiment, it is based on IP The message/data packet of address is transmitted using at least one of predefine/being pre-configured and being pre-stored in processing module routing, should Routing follows TCP transfer mechanisms.
Compared with prior art, main advantages of the present invention are that network performance is more preferable, because of lightweight virtual machine in host The expense of communication is seldom.In addition, multi-host network is simplified, because local crossbar switch is forwarded the messages to using domain socket Gateway crossbar switch is encapsulated without additional message.In addition, implementing the present invention without networks such as IPTable rules, port mappings Configuration of territory.
In one embodiment, since software cross switch or application crossbar switch or software definition are switched to lightweight void The application run in quasi- machine provides network abstraction, so the present invention realizes technological progress, simplifies application and development/dispose and change It has been apt to network performance.
In one embodiment, the present invention can run lightweight virtual machine and high-performance/a scalable network is needed to need It is realized in any application asked, these applications can use this method.For old-fashioned application, API need to be transmitted to crossbar switch Library.New application can be transmitted API and be based on crossbar switch library.
Those skilled in the art are it is understood that can be used any of or new algorithm to implement the present invention.However, it is necessary to It is noted that regardless of which kind of known or new algorithm, present invention to provide a kind of method that can be used in backup operation using, To realize benefit mentioned above and technological progress.
Those of ordinary skill in the art, can it will be appreciated that example in conjunction with described in embodiment disclosed in this specification To realize unit and algorithm steps by the combination of electronic hardware or computer software and electronic hardware.Function be by hardware or Specific application and design constraint depending on technical solution is executed by software.Distinct methods can be used in those skilled in the art Realize the described function of each specific application, but it is not considered that the realization has exceeded the scope of the invention.
Those skilled in the art should be clearly understood that, for convenience and succinct purpose be described, for aforementioned system, device With the detailed operation of unit, the corresponding process in preceding method embodiment is can refer to, details are not described herein again.
In several embodiments provided by the invention, it should be appreciated that disclosed systems, devices and methods can pass through other sides Formula is realized.For example, described device embodiment is only exemplary.For example, dividing elements are only division of logic function, Can be other divisions in practical implementations.For example, multiple units or component can be merged or are integrated into another system, or can Ignore or not execution part feature.In addition, shown or being mutually coupled of discussing or direct-coupling can be realized by some interfaces Or communication connection.INDIRECT COUPLING or communication connection between device or unit can be realized by electronics, machinery or other forms.
When these functions are realized in the form of SFU software functional unit and are sold as independent product or in use, they can Storage is in a computer-readable storage medium.Based on the understanding that technical scheme of the present invention substantially or constitute the prior art Part or the part of technical solution can be realized by the form of software product.Computer software product is stored in storage medium In, including some instructions, it is used to indicate computer node (it can be personal computer, server or network node) and executes this hair All or part of steps of method described in bright embodiment.Above-mentioned storage medium includes:Appointing for program code can be stored What medium, such as USB disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (RandomAccess Memory, RAM), disk or CD.
Unless otherwise expressly provided, the equipment being otherwise in communication with each other need not be continuously in communication with each other.In addition, be in communication with each other Equipment with direct communication or can pass through one or more medium indirect communications.
When individual equipment or article is described herein, it is obvious that more than one equipment/article can be used (whether they are from no can cooperation) replaces the individual equipment/article.Similarly, this document describes more than one equipment or objects In the case of product (no matter whether they can coordinate), it is obvious that it is described more than to replace that individual equipment/article can be used One equipment or article, or equipment/article of different number can be used to replace the equipment or program of shown quantity.Alternatively, setting Standby function and/or feature can not clearly state the miscellaneous equipment with this function/feature to embody by one or more. Therefore, other embodiments of the invention need not include the equipment itself.
Finally, language used in the specification has mainly been selected for readable and directiveness purpose, also may be used The language is selected to describe or limit subject of the present invention.Therefore, the scope of the present invention is not intended to is limited by this detailed description System, but limited by application any claim based on appended by herein.Therefore, the open embodiment of the present invention is intended to illustrate It is bright, and the scope of the invention that unrestricted appended claims illustrate.
About the use of any plural number and/or singular references herein, those skilled in the art can turn from plural form It is changed to singulative, and/or plural form is converted to from singulative, to be suitble to specific environment and/or application.Clearly to rise See, singular/plural form can be expressly recited herein.
Although it is empty using the lightweight of host kernel resources with language verbal description to be directed to structure feature and/or method The embodiment of the system and method for software definition switch between quasi- machine, it is to be understood that, appended claims differ fixed limit In described specific features or method.These specific features and method are as the lightweight virtual machine using host kernel resources Between software definition switch system and method embodiment example and disclose.

Claims (35)

1. the node in a kind of communication system, which is characterized in that the node includes:
Processor;
Memory is coupled to the processor, and the processor is used to execute the multiple modules being present in the memory, The multiple module includes:
At least one interface module, for sending and/or receiving at least one message/data packet, interface is by residing in the section The application initialization of at least one of point;And
At least one processing module, for providing at least one connection to the other nodes of one or more of the communication system Management is to realize that Unified Communication, the connection management are obtained based at least one of following:IP address, shared memory key, Public correspondence port and their arbitrary combination.
2. node according to claim 1, which is characterized in that the interface module is additionally operable to will be associated with the application The message/data packet be discharged at least one queue at least one shared memory.
3. according to the node described in claim 1 and 2, which is characterized in that the processing module is additionally operable to:It extracts in the queue The message/data packet associated with the application, and verify the destination of the message/data packet, the destination For same node, different nodes or its arbitrary combination.
4. node according to claim 3, which is characterized in that if the destination of the message/data packet is same One node, then the processing module be additionally operable in the same node establishment/connected using open field socket.
5. node according to claim 3, which is characterized in that if the destination of the message/data packet is not Same node, then the processing module be additionally operable to the IP address based on the destination and transmit the message/data packet.
6. node according to claim 5, which is characterized in that the message/data packet based on the IP address uses Routing at least one of predefine/is pre-configured and is pre-stored in the processing module to transmit, and the routing follows TCP transmission Mechanism.
7. node according to claim 1, which is characterized in that improve the network performance of container/lightweight virtual machine.
8. node according to claim 1, which is characterized in that provide unified communication interface.
9. node according to claim 1, which is characterized in that improve the performance of the network capabilities of the communication system.
10. node according to claim 1, which is characterized in that the processing module is not for influencing to disappearing in host In the case of ceasing the isolation exchanged, cyberspace is connected using domain socket.
11. node according to claim 1, which is characterized in that IP address associated with the other nodes shared is deposited Reservoir key, public correspondence port or their arbitrary combination are pre-stored/are pre-configured to be in the memory of the node.
12. node according to claim 1, which is characterized in that be lightweight virtual machine preferably.
13. the node in a kind of communication system, which is characterized in that the node includes:
Processor;
Memory is coupled to the processor, and the processor is used to execute the multiple modules being present in the memory, The multiple module includes:
At least one crossbar switch processing module, is used for:
It is received using at least one crossbar switch bank interface by residing at least one of described node application initialization extremely A few message/data packet;
The destination of the message/data packet is verified, the destination is same node, different nodes or its arbitrary combination;
If the destination of the message/data packet is same node, establishment/use opening in the same node Domain socket connection;Or
IP address based on the one or more of the other node received in the message/data packet is by the message/data packet Other nodes are transmitted to, other nodes include lightweight virtual machine process.
14. node according to claim 13, which is characterized in that the message/data packet of the reception is stored at least one In at least one queue in a shared memory.
15. according to the node described in claim 13 and 14, which is characterized in that the crossbar switch processing module is additionally operable to extract Association messages/the data packet being stored in the queue, to verify the destination of the message/data packet.
16. node according to claim 13, which is characterized in that using predefine/being pre-configured and be pre-stored in the friendship Fork closes the routing of at least one of processing module and the message/data packet is transmitted to other nodes, and the routing is abided by Follow TCP transfer mechanisms.
17. node according to claim 13, which is characterized in that the crossbar switch processing module is additionally operable to lead to described Other nodes in letter system establish connection, and other nodes preferably include lightweight virtual machine process, the company It connects based at least one of following and obtains:IP address, shared memory key, public correspondence port and their arbitrary group It closes.
18. node according to claim 13, which is characterized in that improve the network performance of container/lightweight virtual machine.
19. node according to claim 13, which is characterized in that other nodes in the communication system are receiving To after the message/data packet, it is used for:
Message/data packet based on the reception handles new connection request;Or
Message/data packet based on the reception creates at least one data structure;
Update the routing for predefine/being pre-configured and being pre-stored in the crossbar switch processing module, it is preferable that the routing base It is updated by gateway IP, IP mask or its mapping arbitrarily combined in the message/data packet of the reception;
The message/data packet of the reception is scanned to retrieve suitable socket handle;To
If the destination of the message/data packet is same node, establishment/use opening in the same node Domain socket connection;Or
The IP address based on the one or more of other nodes received in the message/data packet disappears described Breath/data packet is transmitted to other nodes.
20. a kind of communication system, which is characterized in that including:
Multiple main frames equipment and multiple server apparatus;
Processor;
At least one crossbar switch of the embedded processor, interconnects with the host equipment and the server apparatus, uses Unified communication interface is provided in the communication between the host equipment and/or the server apparatus, wherein the intersection Switch is used for:
It is received using at least one crossbar switch bank interface by residing at least one of described host application initialization extremely A few message/data packet;
The destination of the message/data packet is verified, the destination is the host, the server or its arbitrary combination;
If the destination of the message/data packet is same node, in same host equipment and/or server apparatus Middle establishment/connected using open field socket;Or
Based on the IP address of at least one other host/server equipment received in the message/data packet, disappear described Breath/data packet is transmitted to other host/server equipment, and other host/server equipment include that lightweight is virtual Machine process.
21. communication system according to claim 20, which is characterized in that improve the network of container/lightweight virtual machine Performance.
22. a kind of method that node by communication system executes, which is characterized in that the method includes:
At least one interface is sent and/or is received by residing in at least one of at least one of described node application initialization Message/data packet;
At least one processing module provides at least one connection management to the other nodes of one or more of the communication system To realize that Unified Communication, the connection management are obtained based at least one of following:It is IP address, shared memory key, public Communication port and their arbitrary combination.
23. according to the method for claim 22, which is characterized in that further include:Associated with the application described it will disappear Breath/data packet is discharged at least one queue at least one shared memory of the node.
24. according to the method described in claim 22 and 23, which is characterized in that further include:Extract in the queue with it is described Using the associated message/data packet;And
The destination of the message/data packet is verified, the destination is same node, different nodes or its arbitrary combination.
25. according to the method for claim 24, which is characterized in that if the destination of the message/data packet is Same node, then the processing module be additionally operable in the same node establishment/connected using open field socket.
26. according to the method for claim 24, which is characterized in that if the destination of the message/data packet is Different nodes, then the processing module be additionally operable to the IP address based on the destination and transmit the message/data packet.
27. according to the method for claim 26, which is characterized in that the message/data packet based on the IP address makes It is transmitted at least one of predefine/being pre-configured and being pre-stored in the processing module routing, the routing follows TCP biographies Send mechanism.
28. according to the method for claim 22, which is characterized in that further include:The processing module is not being influenced to host In the case of the isolation of interior message exchange cyberspace is connected using one or more domain sockets.
29. according to the method for claim 22, which is characterized in that IP address associated with the other nodes, shared Memory key, public correspondence port or their arbitrary combination are pre-stored/are pre-configured to be in the memory of the node.
30. a kind of method that node by communication system executes, which is characterized in that the method includes:
It is received using at least one crossbar switch bank interface by residing at least one of described node application initialization extremely A few message/data packet;
The destination of the message/data packet is verified, the destination is same node, different nodes or its arbitrary combination;
If the destination of the message/data packet is same node, establishment/use opening in the same node Domain socket connection;Or
IP address based on the one or more of the other node received in the message/data packet is by the message/data packet Other nodes are transmitted to, other nodes include lightweight virtual machine process.
31. according to the method for claim 30, which is characterized in that further include:The message/data packet of the reception is stored In at least one queue at least one shared memory.
32. according to the method for claim 31, which is characterized in that further include:
Extraction is stored in association messages/data packet in the queue;To verify the purpose of the message/data packet Ground.
33. according to the method for claim 30, which is characterized in that further include:Using predefine/being pre-configured and be pre-stored in The message/data packet is transmitted to other nodes, the routing by the routing of at least one of crossbar switch processing module Follow TCP transfer mechanisms.
34. according to the method for claim 30, which is characterized in that further include:Establish with it is described in the communication system The connection of other nodes, other nodes preferably include lightweight virtual machine process, and the connection is based on following at least one And obtain:IP address, shared memory key, public correspondence port and their arbitrary combination.
35. according to the method for claim 30, which is characterized in that other nodes in the communication system are connecing After receiving the message/data packet, including:
Message/data packet based on the reception handles new connection request;Or the message/data packet wound based on the reception Build at least one data structure;
Update the routing for predefine/being pre-configured and being pre-stored in crossbar switch processing module, it is preferable that the routing is based on institute The message/data packet for stating reception is updated by gateway IP, IP mask or its mapping arbitrarily combined;
The message/data packet of the reception is scanned to retrieve suitable socket handle;To
If the destination of the message/data packet is same node, establishment/use opening in the same node Domain socket connection;Or
The IP address based on the one or more of other nodes received in the message/data packet disappears described Breath/data packet is transmitted to other nodes.
CN201780009111.0A 2016-05-26 2017-05-22 System and method for software defined switching between lightweight virtual machines using host kernel resources Active CN108604992B (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
IN201641018137 2016-05-26
IN201641018137 2016-05-26
PCT/CN2017/085416 WO2017202272A1 (en) 2016-05-26 2017-05-22 System and method of software defined switches between light weight virtual machines using host kernel resources

Publications (2)

Publication Number Publication Date
CN108604992A true CN108604992A (en) 2018-09-28
CN108604992B CN108604992B (en) 2020-09-29

Family

ID=60412053

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201780009111.0A Active CN108604992B (en) 2016-05-26 2017-05-22 System and method for software defined switching between lightweight virtual machines using host kernel resources

Country Status (2)

Country Link
CN (1) CN108604992B (en)
WO (1) WO2017202272A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110990052A (en) * 2019-11-29 2020-04-10 杭州迪普科技股份有限公司 Configuration saving method and device

Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070277179A1 (en) * 2004-09-29 2007-11-29 Tomoyuki Suzuka Information Processing Apparatus, Communication Processing Method, And Computer Program
CN101436966A (en) * 2008-12-23 2009-05-20 北京航空航天大学 Network monitoring and analysis system under virtual machine circumstance
US7620953B1 (en) * 2004-10-05 2009-11-17 Azul Systems, Inc. System and method for allocating resources of a core space among a plurality of core virtual machines
US20100070552A1 (en) * 2008-09-12 2010-03-18 Charles Austin Parker Providing a Socket Connection between a Java Server and a Host Environment
CN102103518A (en) * 2011-02-23 2011-06-22 运软网络科技(上海)有限公司 System for managing resources in virtual environment and implementation method thereof
CN102520944A (en) * 2011-12-06 2012-06-27 北京航空航天大学 Method for realizing virtualization of Windows application program
KR20130122326A (en) * 2012-04-30 2013-11-07 주식회사 케이티 Lightweight virtual machine image system and method for input/output and generating virtual storage image thereof
CN103503386A (en) * 2012-12-31 2014-01-08 华为技术有限公司 Network device and method for processing message
US20140307548A1 (en) * 2010-01-15 2014-10-16 Dell Products L.P. Information Handling System Data Center Bridging Features with Defined Application Environments
CN104380758A (en) * 2012-04-11 2015-02-25 谷歌公司 Scalable, live transcoding with support for adaptive streaming and failover
CN104520820A (en) * 2012-08-04 2015-04-15 微软公司 Function evaluation using lightweight process snapshots
CN104883302A (en) * 2015-03-18 2015-09-02 华为技术有限公司 Method, device and system for forwarding data packet
CN105407164A (en) * 2011-06-15 2016-03-16 瞻博网络公司 Routing proxy for resource requests and resources
CN105550576A (en) * 2015-12-11 2016-05-04 华为技术服务有限公司 Communication method and device between containers
CN105591815A (en) * 2015-12-10 2016-05-18 北京匡恩网络科技有限责任公司 Network control method for power supply relay device of cloud testing platform

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9329843B2 (en) * 2011-08-02 2016-05-03 International Business Machines Corporation Communication stack for software-hardware co-execution on heterogeneous computing systems with processors and reconfigurable logic (FPGAs)

Patent Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070277179A1 (en) * 2004-09-29 2007-11-29 Tomoyuki Suzuka Information Processing Apparatus, Communication Processing Method, And Computer Program
US7620953B1 (en) * 2004-10-05 2009-11-17 Azul Systems, Inc. System and method for allocating resources of a core space among a plurality of core virtual machines
US20100070552A1 (en) * 2008-09-12 2010-03-18 Charles Austin Parker Providing a Socket Connection between a Java Server and a Host Environment
CN101436966A (en) * 2008-12-23 2009-05-20 北京航空航天大学 Network monitoring and analysis system under virtual machine circumstance
US20140307548A1 (en) * 2010-01-15 2014-10-16 Dell Products L.P. Information Handling System Data Center Bridging Features with Defined Application Environments
CN102103518A (en) * 2011-02-23 2011-06-22 运软网络科技(上海)有限公司 System for managing resources in virtual environment and implementation method thereof
CN105407164A (en) * 2011-06-15 2016-03-16 瞻博网络公司 Routing proxy for resource requests and resources
CN102520944A (en) * 2011-12-06 2012-06-27 北京航空航天大学 Method for realizing virtualization of Windows application program
CN104380758A (en) * 2012-04-11 2015-02-25 谷歌公司 Scalable, live transcoding with support for adaptive streaming and failover
KR20130122326A (en) * 2012-04-30 2013-11-07 주식회사 케이티 Lightweight virtual machine image system and method for input/output and generating virtual storage image thereof
CN104520820A (en) * 2012-08-04 2015-04-15 微软公司 Function evaluation using lightweight process snapshots
CN103503386A (en) * 2012-12-31 2014-01-08 华为技术有限公司 Network device and method for processing message
CN104883302A (en) * 2015-03-18 2015-09-02 华为技术有限公司 Method, device and system for forwarding data packet
CN105591815A (en) * 2015-12-10 2016-05-18 北京匡恩网络科技有限责任公司 Network control method for power supply relay device of cloud testing platform
CN105550576A (en) * 2015-12-11 2016-05-04 华为技术服务有限公司 Communication method and device between containers

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
KHALED Z. IBRAHIM等: ""Characterizing the Performance of Parallel Applications on Multi-socket Virtual Machines"", 《2011 11TH IEEE/ACM INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND GRID COMPUTING》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110990052A (en) * 2019-11-29 2020-04-10 杭州迪普科技股份有限公司 Configuration saving method and device
CN110990052B (en) * 2019-11-29 2023-09-26 杭州迪普科技股份有限公司 Configuration preservation method and device

Also Published As

Publication number Publication date
WO2017202272A1 (en) 2017-11-30
CN108604992B (en) 2020-09-29

Similar Documents

Publication Publication Date Title
US11470001B2 (en) Multi-account gateway
CN104038401B (en) Method and system for interoperability for distributed overlay virtual environments
US8989187B2 (en) Method and system of scaling a cloud computing network
CN103997414B (en) Generate method and the network control unit of configuration information
CN113169928B (en) Logical router including a disaggregated network element
JP2023527999A (en) Loop prevention of virtual L2 network
US10819675B2 (en) Managing network connectivity between cloud computing service endpoints and virtual machines
CN107113298A (en) The method that many leases are supported is provided for RDMA
CN106685787B (en) PowerVM (virtual machine) virtualized network management method and device based on OpenStack
US11516126B2 (en) Techniques for high performant virtual routing capabilities
CN108028813A (en) System and method for using the subnet prefix value being used in the global route header (GRH) of linear forwarding table (LFT) lookup in high-performance computing environment
CN109937400A (en) The stream mode of real-time migration for virtual machine transmits
CN104704778A (en) Method and system for virtual and physical network integration
US20220321566A1 (en) Optimized data-over-cable service interface specifications filter processing for batches of data packets using a single access control list lookup
CN105208053A (en) Method for realizing load balance, device and load balance service system
CN109691026A (en) Method and apparatus for updating multiple multiprotocol label switching (MPLS) two-way converting detection (BFD) sessions
CN110311860A (en) Multi-link load balance method and device under VXLAN
CN103597780A (en) Communication mechanism for multiple interface network nodes
CN109863726A (en) Distributed clients guard station device
CN110431827A (en) Distributed network gate framework is realized using location identifier separated protocol to be used for 3GPP mobility
JP2024503600A (en) Layer 2 networking span ports in virtualized cloud environments
CN107249038A (en) Business datum retransmission method and system
JP2024507146A (en) Packet flow in cloud infrastructure based on cached and non-cached configuration information
JP2024503318A (en) Layer 2 networking using access control lists in virtualized cloud environments
JP2024507143A (en) Scaling IP addresses in overlay networks

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant