CN107995032A

CN107995032A - A kind of method and device that network experimental platform is built based on cloud data center

Info

Publication number: CN107995032A
Application number: CN201711229655.6A
Authority: CN
Inventors: 张娇; 黄韬; 王泽南; 刘江; 潘恬; 杨帆; 刘韵洁
Original assignee: Beijing University of Posts and Telecommunications
Current assignee: Beijing University of Posts and Telecommunications
Priority date: 2017-11-29
Filing date: 2017-11-29
Publication date: 2018-05-04
Anticipated expiration: 2037-11-29
Also published as: CN107995032B

Abstract

An embodiment of the present invention provides a kind of method and device that network experimental platform is built based on cloud data center, belong to field of cloud computer technology, this method includes：Receive the experiment topology request of targeted customer's submission, the experiment topological diagram included according to the experiment topology request, search meets the subgraph of the experiment topological diagram from default cloud data center topological diagram, and the subgraph searched is added in default set；The corresponding subgraph of virtual machine minimum number in calculate node is searched for from the set, is denoted as target subgraph；The virtual machine of the calculate node of the target subgraph is migrated, and the calculate node after virtual machine (vm) migration is converted into the corresponding experiment node of the experiment topological diagram.The present invention implements the method and device provided, reduces the difference with real network environment, and then improves conventional efficient and reduce and spend huge Experiment Platform Construction cost.

Description

A kind of method and device that network experimental platform is built based on cloud data center

Technical field

The invention belongs to field of cloud computer technology, and network experimental platform is built based on cloud data center more particularly to one kind Method and device.

Background technology

With the high speed development of the business such as data mining, net cast, AR/VR, explosion is presented in the flow in data center Formula growth, traditional data center network architecture and agreement are difficult to the demand for supporting new business, so, in the data center New project of network technique and agreement emerges in an endless stream.

However, researcher usually requires to carry out substantial amounts of experiment work in the research process for carrying out data center network Make, put forward algorithm either procotol etc. is verified by really testing, need to build during this large-scale Data center's experiment topology, and the topology for building these forms needs substantial amounts of software and hardware resources, and then need to spend huge Experiment Platform Construction cost, moreover, after the completion of experiment porch is built, the maintenance for the soft hardware equipment in experiment porch Substantial amounts of manpower and materials are needed, when without experiment, these equipment are all usually idle, cause device resource to utilize low.

In view of the above problems, the prior art builds computer network using network simulator NS (Network Simulator) The test platform of network problem experiment, wherein, NS is one and is driven based on the time that expansible, programmable event-driven network is imitated True engine, the modeling and simulation of NS completion networks can be used in user, to carry out the verification and performance test of new algorithm and new agreement Work, in order to support user to be modeled network, NS has pre-defined substantial amounts of network model, such as node, link, queue The network model that can be directly provided Deng, user using NS, can also be expanded on existing network model, be increased new Attribute.User realizes the element of experiment by the program language of NS combination object-orienteds, is imitated using the form of pure software The environment of true experiment, although the network experimental platform that NS combination program languages are built saves greatly than building real network environment The time of amount and cost, but since network element is to model to realize by code, and real network environment is complicated, it is many Special situation is difficult to be stated with model, therefore the network experimental platform built using NS, compared with real network environment, Otherness is larger.

The content of the invention

The purpose of the embodiment of the present invention be to provide it is a kind of based on cloud data center build network experimental platform method and Device, to reduce the difference with real network environment.

Concrete technical scheme is as follows：

A kind of method that network experimental platform is built based on cloud data center, this method are included：

Receive the experiment topology request of targeted customer's submission, wherein, the experiment topology request is to include testing topological diagram Request, it is described experiment topological diagram include the first quantity experiment node and the first quantity test node between link pass System；

The experiment topological diagram included according to the experiment topology request, is searched for full from default cloud data center topological diagram The subgraph of the foot experiment topological diagram, and the subgraph searched is added in default set；

The corresponding subgraph of virtual machine minimum number in calculate node is searched for from the set, is denoted as target subgraph；

The virtual machine of the calculate node of the target subgraph is migrated, and the calculate node after virtual machine (vm) migration is turned It is changed to the corresponding experiment node of the experiment topological diagram.

Further, before the experiment topology request for receiving targeted customer's submission, the method further includes：

The experiment topology request that targeted customer submits is parsed, obtains experiment topological diagram；

When the experiment topological diagram obtained is the subgraph of default cloud data center topological diagram, if in the cloud data The idling-resource of second quantity calculate node of heart topological diagram meets all calculate nodes in the cloud data center topological diagram The resource needed for virtual machine is run, then performs described the step of receiving the experiment topology request that targeted customer submits, wherein, it is described Second quantity is that calculate node total quantity subtracts the quantity that node is tested in experiment topological diagram in cloud data center topological diagram.

Further, it is described from the set search for calculate node on the corresponding subgraph of virtual machine minimum number, including：

The quantity summation of virtual machine in the corresponding calculate node of each subgraph in the set is arranged according to size Sequence, obtains the virtual machine quantity series of each subgraph；

According to the quantity summation of virtual machine in the corresponding calculate node of each subgraph in the virtual machine quantity series Sequence, determines the corresponding subgraph of virtual machine minimum number.

Further, the virtual machine of the calculate node to the target subgraph migrates, and by virtual machine (vm) migration Calculate node afterwards is converted to experiment node, including：

The virtual machine of the calculate node of the target subgraph is migrated using the virtual machine migration method of memory copying；

Calculate node after virtual machine (vm) migration is converted into the corresponding experiment node of the experiment topological diagram.

Further, the calculate node by after virtual machine (vm) migration is converted to the corresponding experiment section of the experiment topological diagram Point, including：

For the calculate node of each required conversion, changed in accordance with the following steps；

Close the cloud calculation service managing process in the calculate node；

After the cloud calculation service managing process is closed, calculating section is closed after preserving the cloud computing kernel of the calculate node Point, wherein, the cloud computing kernel runs the kernel of cloud calculation service for calculate node；

Selection starts from the experiment kernel of the calculate node after restarting the calculate node, and the experiment kernel is calculate node The kernel of operational network experiment.

Further, restart after the calculate node that selection is after the experiment kernel of the calculate node starts described, institute The method of stating further includes：

According to the target experiment user logon account of establishment, the file that the target experiment user produces is recorded.

Further, in the target experiment user logon account according to establishment, the target experiment user production is recorded After raw file, the method further includes：

The file produced according to the target experiment user that is recorded, deletes the target experiment user logon account and described Target experiment user account correspond to produce file, and the experiment kernel to the calculate node initialize after close the calculating Node；

Selection starts from the cloud computing kernel of the calculate node after restarting the calculate node, and opens the industry of the calculate node Business managing process.

Further, migrated in the virtual machine of the calculate node to the target subgraph, and virtual machine is moved After calculate node after shifting is converted to experiment node, the method further includes：

Read the IP network section of the network flow data bag of the interchanger received；

When the IP network section of the network flow data bag is corresponding with the management flow of the cloud computing in default first order flow table IP network section matching when, the network flow data is forwarded a packet to the management flow flow table in default second level flow table, such as The head five-tuple of network flow data bag described in fruit is matched with the five-tuple of the second level flow table, then performs the second level The corresponding action in flow table matching domain respective action domain；Wherein,

The first order flow table includes the communication flows pair between the corresponding IP network section of management flow of cloud computing, virtual machine The corresponding IP network section of IP network section and the flow of network experiment answered, the flow table are to carry out real-time update according to controller, The second level flow table includes management flow flow table, communication flows flow table and empirical flow flow table, and the management flow flow table is Flow table for the management flow for handling cloud computing, the communication flows flow table are logical between cloud computing virtual machine for handling The flow table of letter flow amount, the empirical flow flow table are the flow tables for handling system for cloud computing empirical flow, the second level stream Each flow table in table includes matching domain and action fields corresponding with matching domain, and the matching domain includes five-tuple, described Five-tuple includes source host IP fields, source host peer-port field, destination host IP fields, destination host peer-port field and network association Discuss field；

Communication flows between the virtual machine in the IP network section and default first order flow table of the network flow data bag During corresponding IP network section matching, the network flow data is forwarded a packet to the communication flows flow table of default second level flow table In, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, described in execution The corresponding action in second level flow table matching domain respective action domain；

When the IP network section of the network flow data bag is corresponding with the flow of the network experiment in default first order flow table When IP network section matches, the network flow data is forwarded a packet in the empirical flow flow table of default second level flow table, if The head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, then performs the second level stream The corresponding action in table matching domain respective action domain.

A kind of network experiment device in data based on cloud, described device include：

Request receiving module, for receiving the experiment topology request of targeted customer's submission, wherein, the experiment topology request It is the request for including testing topological diagram, the experiment topological diagram includes the first quantity experiment node and the first quantity experiment section Linking relationship between point；

Subgraph search module, for the experiment topological diagram included according to the experiment topology request, from default cloud data Search meets the subgraph of the experiment topological diagram in centric topology figure, and the subgraph searched is added in default set；

Target subgraph search module, for searching for the corresponding son of virtual machine minimum number in calculate node from the set Figure, is denoted as target subgraph；

Node modular converter is tested, the virtual machine for the calculate node to the target subgraph migrates, and by void Calculate node after the migration of plan machine is converted to the corresponding experiment node of the experiment topological diagram.

Further, described device further includes：

Network flow data bag read module, the IP network section of the network flow data bag for reading the interchanger received；

First matching module, by based on the cloud in the IP network section of the network flow data bag and default first order flow table When the corresponding IP network section of management flow of calculation matches, the network flow data is forwarded a packet in default second level flow table Flow flow table is managed, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, Then perform the corresponding action in the second level flow table matching domain respective action domain；Wherein,

Second matching module, for when the IP network section of the network flow data bag with it is virtual in default first order flow table When the corresponding IP network section of communication flows between machine matches, the network flow data is forwarded a packet into default second level flow table Communication flows flow table in, if the five-tuple of the head five-tuple of the network flow data bag and the second level flow table Match somebody with somebody, then perform the corresponding action in the second level flow table matching domain respective action domain；

3rd matching module, for when the network in the IP network section of the network flow data bag and default first order flow table When the corresponding IP network section of flow of experiment matches, the network flow data is forwarded a packet to the experiment of default second level flow table In flow flow table, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, Perform the corresponding action in the second level flow table matching domain respective action domain.

At the another aspect that the present invention is implemented, a kind of computer-readable recording medium is additionally provided, it is described computer-readable Instruction is stored with storage medium, when run on a computer so that computer performs a kind of any of the above-described base In the method that cloud data center builds network experimental platform.

At the another aspect implemented of the present invention, the embodiment of the present invention additionally provides a kind of electronic equipment, including processor, logical Believe interface, memory and communication bus, wherein, processor, communication interface, memory is completed mutual logical by communication bus Letter；

Memory, for storing computer program；

Processor, during for performing the program stored on memory so that computer performs any of the above-described described one The method that kind builds network experimental platform based on cloud data center.

A kind of method and device that network experimental platform is built based on cloud data center provided in an embodiment of the present invention, can be with The experiment topology request submitted by receiving targeted customer, according to the experiment topological diagram that includes of experiment topology request, from pre- If cloud data center topological diagram in search meet the subgraph of the experiment topological diagram, and the subgraph searched is added to default Set in；The corresponding subgraph of virtual machine minimum number in calculate node is searched for from the set, to the target subgraph The virtual machine of calculate node is migrated, and by the calculate node after virtual machine (vm) migration be converted to it is described experiment topological diagram it is corresponding Test node.This method from the subgraph with testing the matched cloud data center topological diagram of topological diagram by searching in calculate node The minimum subgraph of virtual machine, the corresponding virtual machine of calculate node in the minimum subgraph of virtual machine is migrated, and will migration Calculate node afterwards is converted to the corresponding experiment node of experiment topological diagram, reduces the difference with real network environment, Jin Erti High conventional efficient and reduce and spend huge Experiment Platform Construction cost.Certainly, implement any of the products of the present invention or square Method must be not necessarily required to reach all the above advantage at the same time.

Brief description of the drawings

In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, below will be to embodiment or existing There is attached drawing needed in technology description to be briefly described, it should be apparent that, drawings in the following description are only this Some embodiments of application, for those of ordinary skill in the art, without creative efforts, can be with Other attached drawings are obtained according to these attached drawings.

Fig. 1 be it is provided by the embodiments of the present application the first built based on cloud data center network experimental platform method stream Journey schematic diagram；

Fig. 2 is cloud data center topological diagram provided by the embodiments of the present application；

Fig. 3 is the stream of the second provided by the embodiments of the present application method that network experimental platform is built based on cloud data center Journey schematic diagram；

Fig. 4 be it is provided by the embodiments of the present application the third built based on cloud data center network experimental platform method stream Journey schematic diagram；

Fig. 5 is the flow diagram that data on flows bag provided by the embodiments of the present application matches flow table；

Fig. 6 is a kind of structure for the device that network experimental platform is built based on cloud data center provided by the embodiments of the present application Schematic diagram；

Fig. 7 is a kind of electronic equipment provided by the embodiments of the present application.

Embodiment

Below in conjunction with the attached drawing in the embodiment of the present application, the technical solution in the embodiment of the present application is carried out clear, complete Site preparation describes, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on Embodiment in the application, those of ordinary skill in the art are obtained every other without creative efforts Embodiment, shall fall in the protection scope of this application.

Fig. 1 is the first method that network experimental platform is built based on cloud data center provided by the embodiments of the present application, should Method includes：

S101, receives the experiment topology request of targeted customer's submission, wherein, the experiment topology request is to include experiment to open up The request of figure is flutterred, the experiment topological diagram includes the link between the first quantity experiment node and the first quantity experiment node Relation；

It should be noted that when carrying out network experiment, it is necessary to select when the calculate node progress to cloud data center is topological Corresponding physical node and physical link are selected, required Experimental Network topology is formed, Topological Mapping will be tested into cloud data In heart network, experiment topological diagram is exactly the topological diagram set in advance, waits to need to be mapped in cloud data center network.

S102, the experiment topological diagram included according to the experiment topology request, from default cloud data center topological diagram Search meets the subgraph of the experiment topological diagram, and the subgraph searched is added in default set；

Wherein, cloud data center topological diagram includes switching node 21 and calculate node 22, and linking relationship therebetween is such as Shown in Fig. 2.

To testing topology in cloud data center network maps, the target of the mapping is in cloud data center topology Selected section calculate node in the calculate node of figure, to meet the topological demand of experiment, i.e. the part calculate node is and experiment The matched subgraph of topological diagram, selects a subgraph to meet experiment topology, as shown in Figure 1.

S103, the corresponding subgraph of virtual machine minimum number in calculate node is searched for from the set, is denoted as target subgraph；

It is noted that because the virtual machine, it is necessary to operating above is selected as the calculate node for testing node Migrated, so when selecting subgraph so that the subgraph of selection can meet the needs of experiment topology, while migrate as far as possible most The virtual machine of small number, in the hope of causing the influence of minimum to cloud calculation service.

Based on it is above-mentioned to experiment topology description in, meet experiment topological demand be from part calculate node i.e. with experiment The corresponding subgraph of virtual machine minimum number is selected in the matched subgraph of topological diagram.

A kind of embodiment for realizing this step is：

The embodiment can efficiently and effectively select the corresponding subgraph of virtual machine minimum number, to realize to cloud computing Business causes the influence of minimum, and then reduces expense.

S104, migrates the virtual machine of the calculate node of the target subgraph, and by the calculating after virtual machine (vm) migration Node is converted to the corresponding experiment node of the experiment topological diagram.

It is noted that virtual machine is being selected at least after corresponding subgraph, it is necessary to which selected subgraph is corresponding Virtual machine in calculate node is migrated, and calculate node of soaring is tested, and the calculate node soared is converted into progress The experiment node of experiment.

A kind of embodiment for implementing this step is：Using the virtual machine migration method of memory copying to the target subgraph The virtual machine of calculate node migrated；

The embodiment is using the virtual machine migration method of memory copying to the virtual of the calculate node of the target subgraph Machine is migrated, and improves the accuracy rate and efficiency of virtual machine (vm) migration.

When calculate node is selected as the experiment node in experiment topological diagram, it is necessary to which the calculate node is converted to experiment Node.Since in an experiment, experiment user may change the kernel setup of system, in order to the operation of experiment do not influence it is follow-up Cloud calculation service, there is two kernels, two kernels of calculate node to be referred to as the operating system of each calculate node Cloud computing kernel and experiment kernel.

Now by taking single calculate node as an example, illustrate that calculate node is converted to the process of experiment node：

Close the cloud calculation service managing process in the calculate node；

Selection starts from the experiment kernel of the calculate node after restarting the calculate node, and the experiment kernel is calculate node The kernel of operational network experiment；

Restart selection after the calculate node described and after the experiment kernel of the calculate node starts, further include：

When experiment terminates, experiment node needs to be converted into calculate node, virtual machine is continued to run with, to support cloud computing industry Business.The process that specific experiment node is converted to calculate node can be：

Before S101, the method further includes：

When the experiment topological diagram obtained is the subgraph of default cloud data center topological diagram, if in the cloud data The idling-resource of second quantity calculate node of the heart meets that all calculate node operations are empty in the cloud data center topological diagram Resource needed for plan machine, then perform described the step of receiving the experiment topology request that targeted customer submits, wherein, the cloud data Centric topology figure includes calculate node and switching node, and second quantity is calculate node sum in cloud data center topological diagram Amount subtracts the quantity that node is tested in experiment topological diagram.

After the S104, the method further includes：

It is noted that this step can be placed in before step S101, it can also be placed in after S104, the present invention is implemented Example does not limit this.

Method provided in an embodiment of the present invention by from test the matched cloud data center topological diagram of topological diagram subgraph The minimum subgraph of virtual machine in middle search calculate node, by the corresponding virtual machine of calculate node in the minimum subgraph of virtual machine into Row migration, and the calculate node after migration is converted into the corresponding experiment node of experiment topological diagram, caused to cloud calculation service On the basis of minimum influence, the difference with real network environment is reduced, and then improves conventional efficient and reduces cost Huge Experiment Platform Construction cost.

Fig. 3 is the second provided by the embodiments of the present application method that network experimental platform is built based on cloud data center, should Method includes：

S201, the experiment topology request that parsing targeted customer submits, obtains experiment topological diagram；

Targeted customer tests topology request needed for need to submitting, please to experiment topology according to the experiment topology request of submission Ask and parsed, obtain experiment topological diagram, that is to say, that experiment topology request includes testing topological diagram.

S202, when the experiment topological diagram obtained is the subgraph of default cloud data center topological diagram, if the cloud The idling-resource of second quantity calculate node of data center meets all calculate node operation virtual machines in cloud data center Required resource, then perform S203, wherein, the second quantity is that calculate node total quantity subtracts experiment in cloud data center topological diagram The quantity of node is tested in topological diagram.

Whether the experiment topological diagram based on above-mentioned acquisition, judgment experiment topological diagram are corresponding in cloud data center topological diagram Subgraph, if not the subgraph of cloud data center topological diagram, then illustrates that current cloud data center topological diagram can not meet that experiment is opened up The demand flutterred, will refuse the experiment topology request.

In addition, when it is corresponding subgraph in cloud data center topological diagram to test topological diagram, the meter in corresponding subgraph is judged Whether operator node resource is enough, it is assumed that total calculate node number is N, and the experiment number of nodes for testing needs is n, then remaining (N- N) can a calculate node go to carry the virtual machine being currently running in N number of calculate node, if cannot carry, due to virtual machine The cloud calculation service of middle operation can be affected, so refusing the experiment topology request.If resource is enough, experiment topology please Ask and received, implement that the process of topology request mapping will be tested.

S203, receives the experiment topology request of targeted customer's submission, wherein, the experiment topology request is to include experiment to open up The request of figure is flutterred, the experiment topological diagram includes the link between the first quantity experiment node and the first quantity experiment node Relation；

Wherein, step S203-S204 is identical with the method that step S101~S102 in Fig. 1 embodiments is performed respectively.Cause This, all embodiments in Fig. 1 are suitable for Fig. 3, and can reach the same or similar beneficial effect, and details are not described herein.

S204, the experiment topological diagram included according to the experiment topology request, from default cloud data center topological diagram Search meets the subgraph of the experiment topological diagram, and the subgraph searched is added in default set；

S205, to the quantity summation of virtual machine in the corresponding calculate node of each subgraph in the set according to size into Row sequence, obtains the virtual machine quantity series of each subgraph；

Wherein, above-mentioned be ranked up according to size can select descending sequence, can also select ascending sequence.

S206, according to the quantity summation of virtual machine in the corresponding calculate node of each subgraph in the virtual machine quantity series In sequence, determine the corresponding subgraph of virtual machine minimum number, be denoted as target subgraph；

In this step, it is ranked up if descending, then selects last corresponding subgraph of virtual machine quantity summation, It is ranked up if ascending, then the corresponding subgraph of first virtual machine quantity summation.

S207, migrates the virtual machine of the calculate node of the target subgraph, and by the calculating after virtual machine (vm) migration Node is converted to the corresponding experiment node of the experiment topological diagram.

Wherein, step S207 is identical with the method that the step S104 in Fig. 1 embodiments is performed respectively.Therefore, the institute in Fig. 1 There is embodiment to be suitable for Fig. 3, and can reach the same or similar beneficial effect, details are not described herein.

Whether method provided in an embodiment of the present invention is in cloud data center topological diagram by using judgment experiment topological diagram Whether corresponding subgraph and the idling-resource for the second quantity calculate node for determining whether cloud data center topological diagram meet Resource in cloud data center needed for all calculate node operation virtual machines of topological diagram tests topology request to determine whether to receive, So that it is more intelligent, the experience effect of user is added, from the son with testing the matched cloud data center topological diagram of topological diagram The subgraph that virtual machine is minimum in calculate node is searched in figure, by the corresponding virtual machine of calculate node in the minimum subgraph of virtual machine Migrated, and the calculate node after migration is converted into the corresponding experiment node of experiment topological diagram, made to cloud calculation service Into on the basis of minimum influence, the difference with real network environment is reduced, and then improves conventional efficient and reduces flower Take huge Experiment Platform Construction cost.

Fig. 4 is the method provided by the embodiments of the present application that the third builds network experimental platform based on cloud data center, should Method includes：

S301, the experiment topology request that parsing targeted customer submits, obtains experiment topological diagram；

Wherein, step S301-S306 is identical with the method that step S201~S206 in Fig. 3 embodiments is performed respectively.Cause This, all embodiments in Fig. 3 are suitable for Fig. 4, and can reach the same or similar beneficial effect, and details are not described herein.

S302, when the experiment topological diagram obtained is the subgraph of default cloud data center topological diagram, if the cloud The idling-resource of second quantity calculate node of data center topology figure meets all meters in the cloud data center topological diagram Resource needed for operator node operation virtual machine, then perform S303, wherein, the second quantity is that section is calculated in cloud data center topological diagram Point total quantity subtracts the quantity that node is tested in experiment topological diagram.

S303, receives the experiment topology request of targeted customer's submission, wherein, the experiment topology request is to include experiment to open up The request of figure is flutterred, the experiment topological diagram includes the link between the first quantity experiment node and the first quantity experiment node Relation；

S304, the experiment topological diagram included according to the experiment topology request, from default cloud data center topological diagram Search meets the subgraph of the experiment topological diagram, and the subgraph searched is added in default set；

S305, to the quantity summation of virtual machine in the corresponding calculate node of each subgraph in the set according to size into Row sequence, obtains the virtual machine quantity series of each subgraph；

S306, according to the quantity summation of virtual machine in the corresponding calculate node of each subgraph in the virtual machine quantity series In sequence, determine the corresponding subgraph of virtual machine minimum number, be denoted as target subgraph；

S307, carries out the virtual machine of the calculate node of the target subgraph using the virtual machine migration method of memory copying Migration；

It is noted that using the virtual machine migration method of memory copying to the void of the calculate node of the target subgraph Plan machine migrated only one of them can realize the embodiment of virtual machine (vm) migration, and invention does not limit this.

S308, for the calculate node of each required conversion, is changed according to step A~step F；

Step A, closes the cloud calculation service managing process in the calculate node；

Wherein, service management process can be understood as being responsible for the management of cloud calculation service, including receive cloud calculation service use The request at family, performs the function such as user's request, automation operation management cloud computing platform.

Step B, after the cloud calculation service managing process is closed, is closed after preserving the cloud computing kernel of the calculate node Calculate node, wherein, the cloud computing kernel runs the kernel of cloud calculation service for calculate node；

Wherein, kernel is the core of an operating system.It is that hardware based first layer software expands, there is provided operating system Most basic function, be operating system work basis, it is responsible for the process of system, memory, device driver, text Part and network system, decide the performance and stability of system, it is therefore desirable to which cloud computing kernel is preserved.

Step C, from the experiment kernel startup of the calculate node, the experiment kernel is meter for selection after restarting the calculate node The kernel of operator node operational network experiment；

Start from the experiment kernel of the calculate node and calculate node namely be converted into experiment node.

Step D, according to the target experiment user logon account of establishment, records the file that the target experiment user produces；

Need to be that target experiment user creates system login account, and then target experiment user after experiment node is converted to It can be tested using the experiment node, and then the file for doing experimental data can be produced, it is therefore desirable to which testing node needs this article Part is recorded.

Step E, the file produced according to the target experiment user recorded, deletes the target experiment user logon account And the target experiment user account corresponds to the file produced, and the experiment kernel to the calculate node initialize after close The calculate node；

It should be noted that when experiment terminates, experiment node needs to be converted into calculate node, continues to run with virtual machine, To support cloud calculation service.

Step F, selection starts from the cloud computing kernel of the calculate node after restarting the calculate node, and opens the calculating section The service management process of point.

It should be noted that it is exactly to test node to converted calculate node to start from the cloud computing kernel of the calculate node.

S309, reads the IP network section of the network flow data bag of the interchanger of reception；

Read receive interchanger network flow data bag IP network section, check the network segment whether with first order flow table The corresponding IP network section of management flow of cloud computing, the corresponding IP network section of communication flows between virtual machine and network experiment Whether the corresponding IP network section of flow matches.

It should be noted that can use software definition, that is, software defined network (Software Defined Network, SDN mode) is managed different types of flow, and whole interchangers in cloud data center, which belong to, supports network exchange mould The SDN switch of type OpenFlow, they are all uniformly controlled by a controller, and controller instructs to hand over by OpenFlow agreements Change planes and carry out the forwarding of flow.

S310, when the IP network section and the management flow of the cloud computing in default first order flow table of the network flow data bag During corresponding IP network section matching, the network flow data is forwarded a packet into the management traffic flow in default second level flow table Table, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, described in execution The corresponding action in second level flow table matching domain respective action domain；Wherein,

It should be noted that cloud calculation service and the business of network experiment not only are carry at the same time in cloud data center, also At the same time there is three classes flow, which is the communication flows and net between the management flow of cloud computing, virtual machine respectively The flow of network experiment, the three classes flow share same set of switching equipment, so need to isolate flow on switching equipment, Processing is forwarded to respectively.

In addition, corresponding act of above-mentioned execution can be understood as forwarding from some switch port.

It is noted that each flow table in first order flow table and second level flow table includes matching domain and and matching domain Corresponding action fields, wherein, the communication stream managed between the corresponding IP network section of flow, virtual machine of the cloud computing of first order flow table The corresponding IP network section of flow for measuring corresponding IP network section and network experiment is exactly matching domain, each flow table in the flow table of the second level Exactly manage flow flow table, communication flows flow table or empirical flow flow table, the matching domain of each flow table is five in the flow table of the second level Tuple.

It is clearer in order to what is described to flow table, below to its specific detailed description：

The type of flow is divided using the network segment, the division of the network segment is as shown in table 1.For example, the stream of the 10.0.0.0/24 network segments Amount belongs to communication flows between virtual machine, and the communication flows between virtual machine between calculate node is all by being assigned with the net The network interface of section is transmitted, and other two classes flow is transmitted by being assigned with the network interface of other network segments.So calculate node Server needs three network interfaces, and different types of flow is transmitted by different network interfaces, add between flow every From property.

Table 1

	The management flow of cloud computing	Communication flows between virtual machine	The flow of network experiment
				The network segment (IP sections)	178.0.0.0/24	10.0.0.0/24	192.0.0.0/24

In SDN switch, in order to which the forwarding to flow is accomplished effectively to isolate, and OpenFlow flow is effectively managed Table, the flow table for belonging to different type flow is placed in different tables.In order to reduce the quantity of flow table, using two-stage flow table, Level-one flow table is placed in single table, the network segment of matched data bag, and the network segment according to belonging to data packet forwards a packet to data Handled in different tables, the flow of data on flows bag matching flow table is as shown in figure 5, therefore first order flow table includes cloud computing The corresponding IP network section of management flow, the flow of the corresponding IP network section of communication flows between virtual machine and network experiment corresponds to IP network section, matching domain and action fields corresponding with matching domain, the second level flow table include Fig. 5 in 1~table of table 3, its In, table 1 is management flow flow table, and table 2 is communication flows flow table, and table 3 is empirical flow flow table, such as belongs to the number of empirical flow Interchanger is reached according to bag, then is matched first in first order flow table, that is, table 0, according to its network segment, matches first order flow table Flow i.e. in table 0 performs " send to Table 3 " i.e. " the transmission action of table 3 ", flow progress data packet being sent in table 3 How matching, the flow table of table 3 are forwarded to according to the specific features reselection of data packet.

S311, the communication between the virtual machine in the IP network section and default first order flow table of the network flow data bag When the corresponding IP network section of flow matches, the network flow data is forwarded a packet to the communication flows stream of default second level flow table In table, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, institute is performed State the corresponding action in second level flow table matching domain respective action domain；

Such as the above-mentioned description to communication flows flow table, network flow data is being forwarded a packet into default second level flow table When in the table 2 of Fig. 5, when the head five-tuple of network flow data bag is matched with the five-tuple of second level flow table, then perform described in The corresponding action in second level flow table matching domain respective action domain.

S312, when the IP network section and the flow pair of the network experiment in default first order flow table of the network flow data bag During the IP network section matching answered, the network flow data is forwarded a packet in the empirical flow flow table of default second level flow table, If the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, described second is performed The corresponding action in level flow table matching domain respective action domain.

Such as the above-mentioned description to communication flows flow table, network flow data is being forwarded a packet into default second level flow table When in the table 3 of Fig. 5, when the head five-tuple of network flow data bag is matched with the five-tuple of second level flow table, then perform described in The corresponding action in second level flow table matching domain respective action domain.

Whether method provided in an embodiment of the present invention is in cloud data center topological diagram by using judgment experiment topological diagram Whether corresponding subgraph and the idling-resource for the second quantity calculate node for determining whether cloud data center topological diagram meet Resource in cloud data center in topological diagram needed for all calculate node operation virtual machines is asked to determine whether to receive experiment topology Ask so that it is more intelligent, the experience effect of user is added, from testing the matched cloud data center topological diagram of topological diagram The subgraph that virtual machine is minimum in calculate node is searched in subgraph, the calculate node in the minimum subgraph of virtual machine is corresponding virtual Machine is migrated, and the calculate node after migration is converted to the corresponding experiment node of experiment topological diagram, to cloud calculation service On the basis of the influence for causing minimum, the difference with real network environment is reduced, and then improves conventional efficient and reduces Huge Experiment Platform Construction cost is spent, after experiment porch is put up, when needing experiment, calculate node is converted into reality Node is tested, at the end of experiment, experiment node is converted to calculate node, continues to run with virtual machine, to support cloud calculation service, into One step adds user experience effect；In addition, it is real physical rings that the embodiment of the present invention, which provides the network experimental platform built, Border, is not introduced into the technology of virtualization, and host is physical server, and link is physical link, and forwarding unit is physical switches, this Sample ensure that the authenticity of experiment, accuracy as a result.

Build based on cloud data center that network experimental platform method is corresponding, and the embodiment of the present application additionally provides one with above-mentioned Kind builds the device of network experimental platform based on cloud data center.

Fig. 6 is a kind of device that network experimental platform is built based on cloud data center provided by the embodiments of the present application, the dress Put including：

Request receiving module 401, for receiving the experiment topology request of targeted customer's submission, wherein, the experiment topology Request is the request for including testing topological diagram, and the experiment topological diagram includes the first quantity experiment node and the first quantity is real Test the linking relationship between node；

Subgraph search module 402, for the experiment topological diagram included according to the experiment topology request, from default cloud number Meet the subgraph of the experiment topological diagram according to search in centric topology figure, and the subgraph searched is added to default set In；

Target subgraph search module 403, corresponds to for searching for virtual machine minimum number in calculate node from the set Subgraph, be denoted as target subgraph；

Node modular converter 404 is tested, the virtual machine for the calculate node to the target subgraph migrates, and will Calculate node after virtual machine (vm) migration is converted to the corresponding experiment node of the experiment topological diagram.

Wherein, described device can also include：

Request analysis module, for parsing the experiment topology request of targeted customer's submission, obtains experiment topological diagram；

Execution module is asked, for working as the subgraph that obtained experiment topological diagram is default cloud data center topological diagram When, if the idling-resource of the second quantity calculate node of the cloud data center topological diagram meets that the cloud data center is opened up The resource needed for all calculate node operation virtual machines in figure is flutterred, then performing the experiment topology for receiving targeted customer's submission please The step of asking, wherein, second quantity is that calculate node total quantity is subtracted in experiment topological diagram in cloud data center topological diagram Test the quantity of node.

The target subgraph search module 403 includes：

Sorting sub-module, for the quantity summation to virtual machine in the corresponding calculate node of each subgraph in the set It is ranked up according to size, obtains the virtual machine quantity series of each subgraph；

Determination sub-module, for the quantity summation according to virtual machine in the corresponding calculate node of each subgraph described virtual Sequence in machine quantity series, determines the corresponding subgraph of virtual machine minimum number.

The experiment node modular converter 404 includes：

Submodule is migrated, for the virtual machine migration method using memory copying to the calculate node of the target subgraph Virtual machine is migrated；

Transform subblock, for the calculate node after virtual machine (vm) migration to be converted to the corresponding experiment of the experiment topological diagram Node.

The transform subblock includes：

Calculate node execution unit, for the calculate node for each required conversion, is changed in accordance with the following steps；

Close the service management process in the calculate node；

After the service management process is closed, calculate node is closed after preserving the cloud computing kernel of the calculate node, its In, the cloud computing kernel runs the kernel of cloud calculation service for calculate node；

Selection starts from the experiment kernel of the calculate node after restarting the calculate node, and the experiment kernel is calculating The kernel of node operational network experiment.

The calculate node execution unit can also include：

Described device can also include：

Device provided in an embodiment of the present invention by from test the matched cloud data center topological diagram of topological diagram subgraph The minimum subgraph of virtual machine in middle search calculate node, by the corresponding virtual machine of calculate node in the minimum subgraph of virtual machine into Row migration, and the calculate node after migration is converted into the corresponding experiment node of experiment topological diagram, caused to cloud calculation service On the basis of minimum influence, the difference with real network environment is reduced, and then improves conventional efficient and reduces cost Huge Experiment Platform Construction cost.

The embodiment of the present invention additionally provides a kind of electronic equipment, as shown in fig. 7, comprises processor 501, communication interface 502, Memory 503 and communication bus 504, wherein, processor 501, communication interface 502, memory 503 is complete by communication bus 504 Into mutual communication,

Memory 503, for storing computer program；

Processor 501, during for performing the program stored on memory 503, realizes provided in an embodiment of the present invention one The method that kind builds network experimental platform based on cloud data center.

Specifically, a kind of above-mentioned method that network experimental platform is built based on cloud data center, including：

It can be seen from the above that electronic equipment provided in this embodiment is performed, by from the experiment matched cloud data of topological diagram The subgraph that virtual machine is minimum in calculate node is searched in the subgraph of heart topological diagram, by the calculate node in the minimum subgraph of virtual machine Corresponding virtual machine is migrated, and the calculate node after migration is converted to the corresponding experiment node of experiment topological diagram, right On the basis of cloud calculation service causes the influence of minimum, the difference with real network environment is reduced, and then improves experiment effect The rate Experiment Platform Construction cost huge with cost is reduced.

Above-mentioned related content builds the embodiment of network experimental platform based on cloud data center and preceding method is implemented What example part provided builds that network experimental platform determination mode is identical, and which is not described herein again based on cloud data center.

The communication bus that above-mentioned electronic equipment is mentioned can be Peripheral Component Interconnect standard (Peripheral Component Interconnect, PCI) bus or expanding the industrial standard structure (Extended Industry Standard Architecture, EISA) bus etc..The communication bus can be divided into address bus, data/address bus, controlling bus etc..For just Only represented in expression, figure with a thick line, it is not intended that an only bus or a type of bus.

Communication interface is used for the communication between above-mentioned electronic equipment and other equipment.

Memory can include random access memory (Random Access Memory, RAM), can also include non-easy The property lost memory (Non-Volatile Memory, NVM), for example, at least a magnetic disk storage.Optionally, memory may be used also To be at least one storage device for being located remotely from aforementioned processor.

Above-mentioned processor can be general processor, including central processing unit (Central Processing Unit, CPU), network processing unit (Network Processor, NP) etc.；It can also be digital signal processor (Digital Signal Processing, DSP), it is application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), existing It is field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete Door or transistor logic, discrete hardware components.

In another embodiment provided by the invention, a kind of computer-readable recording medium is additionally provided, which can Read to be stored with instruction in storage medium, when run on a computer so that computer performs any institute in above-described embodiment That states builds network experimental platform method based on cloud data center.

It can be seen from the above that when performing the application program stored in computer-readable recording medium provided in this embodiment, pass through The subgraph that virtual machine is minimum in calculate node is searched for from the subgraph with testing the matched cloud data center topological diagram of topological diagram, will The corresponding virtual machine of calculate node in the minimum subgraph of virtual machine is migrated, and the calculate node after migration is converted to reality The corresponding experiment node of topological diagram is tested, on the basis of the influence of minimum is caused to cloud calculation service, is reduced and live network The difference of environment, and then improve conventional efficient and reduce and spend huge Experiment Platform Construction cost.

In the above-described embodiments, can come wholly or partly by software, hardware, firmware or its any combination real It is existing.When implemented in software, can realize in the form of a computer program product whole or in part.The computer program Product includes one or more computer instructions.When loading on computers and performing the computer program instructions, all or Partly produce according to the flow or function described in the embodiment of the present invention.The computer can be all-purpose computer, special meter Calculation machine, computer network or other programmable devices.The computer instruction can be stored in computer-readable recording medium In, or the transmission from a computer-readable recording medium to another computer-readable recording medium, for example, the computer Instruction can pass through wired (such as coaxial cable, optical fiber, numeral from a web-site, computer, server or data center User's line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or Data center is transmitted.The computer-readable recording medium can be any usable medium that computer can access or It is the data storage devices such as server, the data center integrated comprising one or more usable mediums.The usable medium can be with It is magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disc Solid State Disk (SSD)) etc..

It should be noted that herein, relational terms such as first and second and the like are used merely to a reality Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation In any this actual relation or order.Moreover, term " comprising ", "comprising" or its any other variant are intended to Non-exclusive inclusion, so that process, method, article or equipment including a series of elements not only will including those Element, but also including other elements that are not explicitly listed, or further include as this process, method, article or equipment Intrinsic key element.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that Also there are other identical element in process, method, article or equipment including the key element.

Each embodiment in this specification is described using relevant mode, identical similar portion between each embodiment Divide mutually referring to what each embodiment stressed is the difference with other embodiment.It is real especially for system For applying example, since it is substantially similar to embodiment of the method, so description is fairly simple, related part is referring to embodiment of the method Part explanation.

The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all Any modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present invention It is interior.

Claims

A kind of 1. method that network experimental platform is built based on cloud data center, it is characterised in that the described method includes：

Receive the experiment topology request of targeted customer's submission, wherein, the experiment topology request is to include testing asking for topological diagram Ask, the experiment topological diagram includes the linking relationship between the first quantity experiment node and the first quantity experiment node；

The experiment topological diagram included according to the experiment topology request, searches for from default cloud data center topological diagram and meets institute The subgraph of experiment topological diagram is stated, and the subgraph searched is added in default set；

The corresponding subgraph of virtual machine minimum number in calculate node is searched for from the set, is denoted as target subgraph；

The virtual machine of the calculate node of the target subgraph is migrated, and the calculate node after virtual machine (vm) migration is converted to It is described to test the corresponding experiment node of topological diagram.
2. the method as described in claim 1, it is characterised in that it is described receive targeted customer submission experiment topology request it Before, the method further includes：

The experiment topology request that targeted customer submits is parsed, obtains experiment topological diagram；

When the experiment topological diagram obtained is the subgraph of default cloud data center topological diagram, if the cloud data center is opened up The idling-resource for flutterring the second quantity calculate node of figure meets all calculate node operations in the cloud data center topological diagram Resource needed for virtual machine, then perform described the step of receiving the experiment topology request that targeted customer submits, wherein, described second Quantity is that calculate node total quantity subtracts the quantity that node is tested in experiment topological diagram in cloud data center topological diagram.
3. the method as described in claim 1, it is characterised in that described that virtual machine number in calculate node is searched for from the set The minimum corresponding subgraph of amount, including：

The quantity summation of virtual machine in the corresponding calculate node of each subgraph in the set is ranked up according to size, is obtained To the virtual machine quantity series of each subgraph；

According to sequence of the quantity summation of virtual machine in the corresponding calculate node of each subgraph in the virtual machine quantity series, Determine the corresponding subgraph of virtual machine minimum number.
4. the method as described in claim 1, it is characterised in that the virtual machine of the calculate node to the target subgraph into Row migration, and the calculate node after virtual machine (vm) migration is converted into experiment node, including：

The virtual machine of the calculate node of the target subgraph is migrated using the virtual machine migration method of memory copying；

Calculate node after virtual machine (vm) migration is converted into the corresponding experiment node of the experiment topological diagram.
5. method as claimed in claim 4, it is characterised in that the calculate node by after virtual machine (vm) migration is converted to described The corresponding experiment node of topological diagram is tested, including：

For the calculate node of each required conversion, changed in accordance with the following steps；

Close the cloud calculation service managing process in the calculate node；

After the cloud calculation service managing process is closed, calculate node is closed after preserving the cloud computing kernel of the calculate node, Wherein, the cloud computing kernel runs the kernel of cloud calculation service for calculate node；

Selection starts from the experiment kernel of the calculate node after restarting the calculate node, and the experiment kernel runs for calculate node The kernel of network experiment.
6. method as claimed in claim 5, it is characterised in that restart after the calculate node selection from the calculate node described Experiment kernel start after, the method further includes：

According to the target experiment user logon account of establishment, the file that the target experiment user produces is recorded.
7. method as claimed in claim 6, it is characterised in that in the target experiment user logon account according to establishment, After recording the file that the target experiment user produces, the method further includes：

The file produced according to the target experiment user recorded, deletes the target experiment user logon account and the target Experiment user account correspond to produce file, and the experiment kernel to the calculate node initialize after close the calculating section Point；

Selection starts from the cloud computing kernel of the calculate node after restarting the calculate node, and opens the business pipe of the calculate node Reason process.
8. the method as described in any in claim 1-7, it is characterised in that in the calculate node to the target subgraph Virtual machine migrated, and by the calculate node after virtual machine (vm) migration be converted to experiment node after, the method further includes：

Read the IP network section of the network flow data bag of the interchanger received；

As the IP network section of network flow data bag IP corresponding with the management flow of the cloud computing in default first order flow table When the network segment matches, the network flow data is forwarded a packet to the management flow flow table in default second level flow table, if institute The head five-tuple for stating network flow data bag is matched with the five-tuple of the second level flow table, then performs the second level flow table The corresponding action in matching domain respective action domain；Wherein,

The communication flows that the first order flow table is included between the corresponding IP network section of management flow of cloud computing, virtual machine is corresponding The corresponding IP network section of the flow of IP network section and network experiment, the second level flow table include empirical flow flow table, management flow Flow table, communication flows flow table, the management flow flow table are used for the flow table for the management flow for handling cloud computing, the communication flows Flow table is used for the flow table for handling the communication flows between cloud computing virtual machine, and the empirical flow flow table is used to handle cloud computing net Flow table described in the flow table of network empirical flow is to carry out real-time update according to controller, each flow table in the second level flow table Include matching domain and action fields corresponding with matching domain, the matching domain includes five-tuple, and the five-tuple includes source master Machine IP fields, source host peer-port field, destination host IP fields, destination host peer-port field and procotol field；

When the IP network section of the network flow data bag is corresponding with the communication flows between the virtual machine in default first order flow table IP network section matching when, the network flow data is forwarded a packet in the communication flows flow table of default second level flow table, such as The head five-tuple of network flow data bag described in fruit is matched with the five-tuple of the second level flow table, then performs the second level The corresponding action in flow table matching domain respective action domain；

When the IP network section of network flow data bag IP network corresponding with the flow of the network experiment in default first order flow table During section matching, the network flow data is forwarded a packet in the empirical flow flow table of default second level flow table, if described The head five-tuple of network flow data bag is matched with the five-tuple of the second level flow table, then performs the second level flow table With the corresponding action in domain respective action domain.
9. the network experiment device in a kind of data based on cloud, it is characterised in that described device includes：

Request receiving module, for receiving the experiment topology request of targeted customer's submission, wherein, the experiment topology request is bag Include experiment topological diagram request, it is described experiment topological diagram include the first quantity experiment node and the first quantity test node it Between linking relationship；

Subgraph search module, for the experiment topological diagram included according to the experiment topology request, from default cloud data center Search meets the subgraph of the experiment topological diagram in topological diagram, and the subgraph searched is added in default set；

Target subgraph search module, for searching for the corresponding subgraph of virtual machine minimum number in calculate node from the set, It is denoted as target subgraph；

Node modular converter is tested, the virtual machine for the calculate node to the target subgraph migrates, and by virtual machine Calculate node after migration is converted to the corresponding experiment node of the experiment topological diagram.
10. device as claimed in claim 9, it is characterised in that described device further includes：

Network flow data bag read module, the IP network section of the network flow data bag for reading the interchanger received；

First matching module, for when the cloud computing in the IP network section of the network flow data bag and default first order flow table When the corresponding IP network section of management flow matches, the network flow data is forwarded a packet into the management in default second level flow table Flow flow table, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, holds The corresponding action in the row second level flow table matching domain respective action domain；Wherein,

The communication flows that the first order flow table is included between the corresponding IP network section of management flow of cloud computing, virtual machine is corresponding The corresponding IP network section of the flow of IP network section and network experiment, the second level flow table include empirical flow flow table, management flow Flow table, communication flows flow table, the management flow flow table are used for the flow table for the management flow for handling cloud computing, the communication flows Flow table is used for the flow table for handling the communication flows between cloud computing virtual machine, and the empirical flow flow table is used to handle cloud computing net Flow table described in the flow table of network empirical flow is to carry out real-time update according to controller, each flow table in the second level flow table Include matching domain and action fields corresponding with matching domain, the matching domain includes five-tuple, and the five-tuple includes source master Machine IP fields, source host peer-port field, destination host IP fields, destination host peer-port field and procotol field；

Second matching module, for when the virtual machine in the IP network section of the network flow data bag and default first order flow table it Between the matching of communication flows corresponding IP network section when, the network flow data is forwarded a packet into the logical of default second level flow table In letter flow amount flow table, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, Then perform the corresponding action in the second level flow table matching domain respective action domain；

3rd matching module, for when the network experiment in the IP network section of the network flow data bag and default first order flow table The matching of flow corresponding IP network section when, the network flow data is forwarded a packet to the empirical flow of default second level flow table In flow table, if the head five-tuple of the network flow data bag is matched with the five-tuple of the second level flow table, perform The corresponding action in the second level flow table matching domain respective action domain.