WO2020034492A1 - 虚拟网络功能扩容与缩容方法及系统 - Google Patents
虚拟网络功能扩容与缩容方法及系统 Download PDFInfo
- Publication number
- WO2020034492A1 WO2020034492A1 PCT/CN2018/119286 CN2018119286W WO2020034492A1 WO 2020034492 A1 WO2020034492 A1 WO 2020034492A1 CN 2018119286 W CN2018119286 W CN 2018119286W WO 2020034492 A1 WO2020034492 A1 WO 2020034492A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- vnf
- load
- vnfs
- level
- group
- 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.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45595—Network integration; Enabling network access in virtual machine instances
Definitions
- the invention relates to the field of capacity expansion and reduction of virtual network functions, and in particular to a method and system for capacity expansion and reduction of virtual network functions.
- the virtual network function manager When the virtual network function manager detects that the virtual machine VM (Virtual Machine) cannot handle the current business load, it can migrate the virtual network function to a higher-performance virtual machine through hot migration.
- the migration process is transparent to the user. During the migration, the state of all network connections and all applications while the business is running will be maintained, while the migration time will be minimized.
- the content of the virtual machine migration includes the migration of memory data, network connections, file systems, and various resources. It is necessary to ensure the consistency of the status and data of the source and destination virtual machines before and after the migration.
- the key to the virtual machine migration process is the processing of memory state data.
- VMotion and Xen-Motion use a pre-copy method.
- the virtual machine management program identifies the memory pages as dirty pages, and then passes through multiple rounds of memory.
- the copying process migrates the memory state data from the source virtual machine to the destination virtual machine, and is controlled by a bitmap during the migration process. With the execution of the virtual machine migration process, the memory data is gradually transferred to the destination virtual machine. When the amount of memory data to be transferred is lower than a preset threshold, the iteration is stopped, and all the dirty page data generated in the last round is copied. Go to the destination virtual machine, start the destination virtual machine, and stop the source virtual machine.
- OpenStack supports two types of virtual machine migration, namely online migration and block migration. Online migration saves instance information in shared storage, while block migration requires migrating instance memory data and disk file data at the same time.
- OpenStack first performs migration condition audits, including permissions, parameters, destination host memory capacity, source host and destination virtual machine hypervisor versions, and so on. After passing the audit, the libvirt interface is called to migrate the source host to the destination host.
- detection methods such as round robin are used to track the migration status of the virtual machine until the migration is successful.
- the virtual network function runs in the guest operating system Guest OS of the virtual machine VM.
- the virtual network function needs to be expanded and reduced.
- the main method adopted is to virtual network Function allocation is more appropriate for the virtual machine, and the business data being processed by the virtual network function is migrated to the new virtual machine.
- NS Network Service
- Increasing or reducing the service load will cause multiple virtual network functions to expand or shrink.
- the traditional method is only from the perspective of single virtual network function business expansion. To consider the problem, the global optimum is often not achieved.
- an object of the present invention is to provide a method for expanding and reducing a virtual network function, which can realize the expansion and reduction of a virtual network function from a globally optimal perspective.
- a method for expanding and reducing the capacity of a virtual network includes the following steps:
- VNFs determine the degree of correlation between VNFs.
- VNFs included in each expansion group are simultaneously expanded, or VNFs included in each expansion group are simultaneously reduced.
- the historical service load is divided into multiple levels according to the amount of load
- the time series business association model and the level of business load in the current time period calculate the transition probability of the business load from the current level of business load to other levels, and use the level of the business load corresponding to the maximum transition probability as the current time The level of service load for the next time period adjacent to the segment;
- a high-level priority selection algorithm is set in advance, the high-level service load is selected as the service in the next time period adjacent to the current time period Load level; if a low-level priority selection algorithm is set in advance, a service load with a lower level is selected as the service load level of the next time period adjacent to the current time period.
- the method further includes the following steps: selecting the optimal virtual machine from multiple data centers to run the expanded or reduced VNF.
- the optimal virtual machine running VNF after expansion or reduction is selected from multiple data centers.
- the specific process is as follows:
- the application scenario requirements include the shortest service load transmission time, the shortest VNF path, and the highest security level.
- an idle virtual machine with the smallest difference from the best server is selected as the best virtual machine.
- the degree of correlation between VNFs is determined.
- Multiple VNFs that need to be expanded or reduced are grouped according to the degree of association to form multiple expansions.
- Group or reduction group the specific process is:
- step B After judging all the remaining VNFs according to step B, use the final reference VNF group as the expansion group or the reduction group;
- step D Repeat step D until all VNFs that need to be expanded or reduced are processed.
- the high-dimensional spatial distance is a Chebyshev distance.
- an object of the present invention is to provide a virtual network function expansion and reduction system capable of realizing expansion and reduction of virtual network functions from a globally optimal perspective.
- a virtual network function expansion and reduction system includes:
- Virtual orchestrator which is used to collect historical service load data and analyze it, predict the service load in the future time period, and determine whether to expand or reduce capacity based on the prediction results;
- a virtualization function manager which is used to determine the degree of association between VNFs based on the high-dimensional spatial distance between any two virtual network function VNFs, and group multiple VNFs that need to be expanded or reduced according to the degree of association. Form multiple expansion groups or reduction groups, and expand the VNFs included in each expansion group at the same time, or reduce the VNFs included in each reduction group at the same time.
- the virtual orchestrator includes a business load manager.
- the business load manager collects historical business data and analyzes it, predicts the business load in the future time period, and judges whether to proceed according to the prediction result. Expansion or shrinkage, the specific process is:
- the historical service load is divided into multiple levels according to the amount of load
- the time series business association model and the level of the business load in the current time period calculate the transition probability of the business load from the current level of business load to other levels, and take the level of the business load corresponding to the maximum transition probability as adjacent to the current time period The level of service load for the next time period;
- the service load manager is preset to use a high-level priority selection algorithm or a low-level priority selection algorithm
- a high-level priority selection algorithm is set in advance, the high-level service load is selected as the service in the next time period adjacent to the current time period Load level; if a low-level priority selection algorithm is set in advance, a service load with a lower level is selected as the service load level of the next time period adjacent to the current time period.
- the virtualization orchestrator is further configured to select the optimal virtual machine from multiple data centers to run the expanded or reduced VNF.
- the virtualization orchestrator selects the best virtual machine from multiple data centers to run the expanded or reduced VNF.
- the specific process is as follows:
- the application scenario requirements include the shortest service load transmission time, the shortest VNF path, and the highest security level.
- an idle virtual machine with the smallest difference from the best server is selected as the best virtual machine.
- the virtualization function manager determines the degree of correlation between VNFs according to the high-dimensional spatial distance between any two VNFs, and multiple VNFs that need to be expanded or reduced in capacity according to the degree of correlation Group them to form multiple expansion groups or reduction groups.
- the specific process is as follows:
- step B After judging all the remaining VNFs according to step B, use the final reference VNF group as the expansion group or the reduction group;
- step D Repeat step D until all VNFs that need to be expanded or reduced are processed.
- the high-dimensional spatial distance is a Chebyshev distance.
- the virtual network function expansion and reduction method of the present invention proposes a dynamic combination method to implement virtual network function migration. It is oriented to network services and analyzes the association relationship of virtual network functions to obtain the optimal expansion and reduction combination set.
- the virtual network functions in the expansion group are expanded or reduced at the same time, thereby realizing the expansion and reduction of virtual network functions from a globally optimal perspective.
- FIG. 1 is a flowchart of a method for expanding and reducing a virtual network function according to an embodiment of the present invention
- step S1 is a flowchart of step S1 in a method for expanding and reducing a virtual network function according to an embodiment of the present invention
- FIG. 3 is a block diagram of a virtual network function expansion and reduction system according to an embodiment of the present invention.
- FIG. 4 is a schematic diagram of selecting an optimal virtual machine according to an embodiment of the present invention.
- FIG. 5 is a schematic diagram of dynamic combined capacity expansion or reduction of virtual network functions according to an embodiment of the present invention.
- this embodiment provides a method for expanding and reducing a virtual network function.
- the method includes the following steps:
- VNFs determine the degree of correlation between VNFs.
- Capacity group which simultaneously expands the VNFs contained in each expansion group, or simultaneously reduces the VNFs contained in each expansion group.
- This embodiment proposes a dynamic combination method to implement virtual network function migration. It is oriented to network services and analyzes the association relationship of virtual network functions to obtain the optimal expansion and reduction combination set. Then, the virtual network functions in the same expansion group are simultaneously Capacity expansion or reduction is implemented to achieve the expansion and reduction of virtual network functions from a globally optimal perspective.
- Embodiment 1 collect historical service data for analysis, predict service load in the future time period, and judge whether to expand or reduce capacity based on the prediction result.
- the specific process is:
- the historical service load is divided into multiple levels according to the size of the load
- the service load is divided into ultra high load UH (Ultra High Level), high load H (High level), common load C (Common level), low load L (Low level), and ultra low load UL ( Ultra Low Level) five levels.
- UH Ultra High Level
- high load H High level
- common load C Common level
- low load L Low level
- ultra low load UL Ultra Low Level
- the number of classifications can be adjusted according to the number of business loads, for example, it is divided into seven or nine levels when the range of load changes is large, and it is adjusted to three levels when the range of load changes is small.
- the business load manager establishes a sequence of future business load classifications.
- This invention uses symbols Indicates this classification sequence, where the superscript R indicates the sequence range Range, and the subscript I indicates the virtual network function identifier Identifier.
- the five classification load sequences of the virtual network function VNF3 in the future time period can be expressed as In other words, the VNF3 service load trend is high load, ultra high load, ultra high load, high load, ultra high load.
- analyze the relationship of business load on the time axis and analyze the similarity of business load in different time segments, including peak, holiday, and burst traffic segments, etc., to obtain a quantified time series business load correlation model.
- the business load correlation model based on historical data counts the probability values at each time series point to obtain a business load transfer model.
- a high-level priority selection algorithm is set in advance, the high-level service load is selected as the service in the next time period adjacent to the current time period Load level; if a low-level priority selection algorithm is set in advance, a service load with a lower level is selected as the service load level of the next time period adjacent to the current time period.
- the next possible service load amount is obtained through the service load selection algorithm. If the transfer probability value can be compared, a larger service load transfer probability value is selected as the next service load amount. For example, with for these two transition probability values, the present invention will select the load corresponding to the probability value of 0.75, that is, the ultra-high load as the next business load. If two transition probability values are equal, such as with The invention uses two selection strategies. High-level priority algorithm for business load will return Corresponding service load, that is, ultra-high service load, and the low-level priority selection algorithm for service load will return Corresponding service load, that is, normal service load.
- the method further includes the following steps: selecting the optimal virtual machine from multiple data centers to run the expanded or reduced VNF.
- the data center refers to an operator's cloud-based network data center. After the operator's network is cloud-transformed, the data center built for the cloud-based network is deployed with a telecommunication cloud platform, virtual network functions, management orchestration platform, etc., and is abbreviated in the present invention. data center.
- the optimal virtual machine running the expanded or reduced VNF is selected from multiple data centers.
- the specific process is as follows:
- the application scenario requirements include the shortest service load transmission time, the shortest VNF path, and the highest security level.
- the selection function f DC will use different strategies, such as the shortest service load transfer time. , The shortest virtual network function, the highest security level, etc., and then select the optimal data center.
- the present invention selects an appropriate server region by using a region selection function f RG .
- each area includes multiple servers, and the server selection function f SRV returns the appropriate server to support intelligent expansion or reduction.
- an idle virtual machine with the smallest difference from the best server is selected as the best virtual machine.
- a suitable virtual machine is selected by using a virtual machine selection function f VM .
- the f VM adopts a minimum virtual core number adaptation method that meets requirements, that is, the number of virtual cores required by a virtual network function and each idle virtual machine. The difference between the provided core numbers returns the idle virtual machine with the smallest difference.
- the purpose of this selection method is to retain the virtual machines with a large number of virtual cores for other applications.
- the current server's three virtual machines VM 1 , VM 2 , and VM 3 have a virtual core number of 12, 10, and 6, and the virtual machine selection function f VM
- the virtual machine VM 3 can only provide 6 virtual cores, which fails to meet the virtual network function requirements, and excludes the virtual machine VM 3 .
- calculate the difference between the virtual cores that the virtual machine VM 1 and virtual machine VM 2 can provide and the number of cores required for the virtual network function, and find that the difference between the virtual machine VM 2 is the smallest. Therefore, return the virtual machine VM 2 , that is, VM 2 f VM (VM1, VM2, VM3).
- the four functions f DC , f RG , f SRV , and f VM are used to select a data center, a specific area of the data center, a specific server in the area, and a specific virtual machine on the server, respectively.
- VM 2 f VM (VM 1 , VM 2 , VM 3 ), and finally select the virtual machine VM 2 to run the expanded virtual network function.
- the degree of correlation between VNFs is determined according to the high-dimensional spatial distance between any two VNFs.
- the degree of association is grouped to form multiple expansion groups or reduction groups. The specific process is as follows:
- step B After judging all the remaining VNFs according to step B, use the final reference VNF group as the expansion group or the reduction group;
- step D Repeat step D until all VNFs that need to be expanded or reduced are processed.
- High-dimensional spatial distances can be calculated using the Manhattan distance, Euclidean distance, Minkowski distance, Chebyshev distance, etc. Chebyshev distance is preferred.
- FIG. 5 illustrates a dynamic combined expansion and reduction of virtual network functions.
- the bottom part of Figure 5 is the server, which uses the virtualization engine to build three types of virtualized resource pools: computing, storage, and networking.
- the virtual network function VNF runs on the virtual machine.
- This embodiment proposes an expansion and contraction group selection algorithm based on the Chebyshev distance, and calculates the Chebyshev distance based on the expansion and contraction time and the virtual connection VL bandwidth. (In order to improve accuracy and avoid inaccurate data caused by dimensional differences of different parameters, you can The normalized or adjusted parameter values are used for calculation.),
- the Chebyshev distance is compared with the threshold set by the virtual orchestrator NFVO, and the virtual network function smaller than the threshold is an expansion group .
- the threshold is calculated by the virtual orchestrator NFVO.
- two-dimensional data (tim, ban) is used to represent a virtual network function VNF expansion and contraction time, and a maximum bandwidth of all VLs associated with the virtual network function.
- VNF-1, VNF-2, VNF-3, and VNF-4 and the maximum bandwidth of the associated VL are represented as (tim1, ban1), (tim2, ban2), respectively, as shown in FIG. 5, (tim3, ban3), (tim4, ban4).
- Calculate the six Chebyshev distances between the two virtual network functions for example, calculate the Chebyshev distance between the virtual network function VNF-1 and the virtual network function VNF-2.
- Schiff distance d (1,2) max (
- the near-capacity group includes two virtual network functions, VNF-1 and VNF-3.
- the network function virtualization orchestrator NFVO and the virtual network function manager VNFM start the capacity expansion process, and at the same time VNF-1, VNF-3, and VNF-4 are expanded, and the entire capacity expansion process is managed.
- this embodiment provides a system for expanding and reducing a virtual network function, including:
- Virtual orchestrator which is used to collect historical service load data and analyze it, predict the service load in the future time period, and determine whether to expand or reduce capacity based on the prediction results;
- a virtualization function manager which is used to determine the degree of association between VNFs based on the high-dimensional spatial distance between any two virtual network function VNFs, and group multiple VNFs that need to be expanded or reduced according to the degree of association. Form multiple expansion groups or reduction groups, and expand the VNFs included in each expansion group at the same time, or reduce the VNFs included in each reduction group at the same time.
- the service load manager is located in the Network Function Virtualization Orchestrator (NFVO), and is connected to each Virtual Network Function Manager (VNFM) to obtain each Virtual Network Function (VNF).
- VNF Virtual Network Function
- the virtual network function manager VNFM uploads only the quantified service load index, and does not need to mirror and upload the load data.
- the service data collection cycle is set by the administrator in the network function virtualization orchestrator NFVO and delivered to VNFM and VNF.
- the network function virtualization environment includes a unified network function virtualization orchestrator NFVO, and then corresponds to multiple data centers, including a network function virtualization infrastructure NFVI (Network Function Virtualization Infrastructure) and a virtualization infrastructure manager VIM (Virtualized Infrastructure Manager).
- NFVI Network Function Virtualization Infrastructure
- VIM Virtualized Infrastructure Manager
- Each data center runs one or more virtual network function groups VNFGrp.
- Each virtual network function group VNFGrp includes a virtual network function manager VNFM, a virtual network element manager EM (Element Manger), and multiple virtual networks. Function VNF.
- a virtual network function group VNFGrp can be constructed based on the business function entities of the same manufacturer.
- FIG. 3 it illustrates an example in which a service load manager connects two VNFMs (that is, VNFMA and VNFMB), where VNFMA manages virtual network functions VNF-A1 and VNF-A2, and VNFMB manages virtual network functions VNF-B1 and VNF-B2.
- the virtual network function group VNFGrpA includes four entities, which are a virtual network function manager VNFMA, a virtual network element management EMA, a virtual network function VNF-A1, and a virtual network function VNF-A2.
- the virtualization function entities of the same manufacturer can form a virtual network function group.
- the lower part of Figure 5 is the network function virtualization infrastructure NFVI and the virtual infrastructure manager VIM, which respectively implement hardware virtualization and management of virtual computing, storage, and network resources.
- Service load traffic is generated from the four virtual network functions (ie, VNF-A1, VNF-A2, VNF-B1, VNF-B2) in Figure 3, and the load traffic is uploaded to the virtual network function manager VNFM at a set period and uploaded To the business load manager.
- the service load manager analyzes the historical service load data according to the specific application scenario, determines the appropriate service load collection cycle, and delivers it to each virtual network function entity.
- the resource consumed by the virtual network function is collected by the virtual infrastructure manager VIM and reported to the network function virtualization orchestrator NFVO, and then handed over to the business load manager for analysis.
- the service load manager collects historical service data and analyzes it, predicts the service load in the future time period, and determines whether to expand or shrink based on the prediction result.
- the specific process is:
- the historical service load is divided into multiple levels according to the amount of load
- the time series business association model and the level of the business load in the current time period calculate the transition probability of the business load from the current level of business load to other levels, and take the level of the business load corresponding to the maximum transition probability as adjacent to the current time period The level of service load for the next time period;
- the service load manager is preset to adopt a high-level priority selection algorithm or a low-level priority selection algorithm
- the high-level service load is selected as the service load of the next time period adjacent to the current time period.
- Level if a low-level priority selection algorithm is set in advance, a service load with a lower level is selected as the service load level of the next time period adjacent to the current time period.
- the virtualization orchestrator is further configured to select the optimal virtual machine from multiple data centers to run the virtual network function VNF after the expansion or reduction.
- the virtualization orchestrator selects the best virtual machine from multiple data centers to run the expanded or reduced VNF.
- the specific process is as follows:
- the application scenario requirements include the shortest service load transmission time, the shortest VNF path, and the highest security level.
- an idle virtual machine with the smallest difference from the best server is selected as the best virtual machine.
- the virtualization function manager determines the degree of association between VNFs according to the high-dimensional spatial distance between any two VNFs, and will need to expand capacity or
- the reduced VNFs are grouped according to the degree of association to form multiple expansion groups or reduction groups. The specific process is as follows:
- step B After judging all the remaining VNFs according to step B, use the final reference VNF group as the expansion group or the reduction group;
- step D Repeat step D until all VNFs that need to be expanded or reduced are processed.
- the high-dimensional space distance can be calculated by using the Manhattan distance, the European distance, the Minkowski distance, and the Chebyshev distance. In this embodiment, the Chebyshev distance is preferred.
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
本发明公开了一种虚拟网络功能扩容与缩容方法及系统,涉及虚拟网络功能扩容与缩容领域,该方法包括以下步骤:采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及根据任意两个虚拟网络功能VNF之间的切比雪夫距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。本发明中的虚拟网络功能扩容与缩容方法能从全局最优的角度实现虚拟网络功能的扩容与缩容。
Description
本发明涉及虚拟网络功能扩容与缩容领域,具体涉及一种虚拟网络功能扩容与缩容方法及系统。
虚拟网络功能管理器检测到虚拟机VM(Virtual Machine)无法处理当前业务负载时,可以通过热迁移的方式将虚拟网络功能迁移至性能更高的虚拟机上,迁移过程对用户是透明的,整个迁移期间将保持业务运行时所有网络连接、所有应用程序的状态,同时尽量减小迁移时间。虚拟机迁移内容包括内存数据、网络连接、文件系统、各种资源的迁移,需要保证迁移前后源虚拟机与目的虚拟机状态与数据的一致性。
虚拟机迁移过程中的关键是对内存状态数据的处理,其中VMotion与Xen-Motion采用预拷贝方法,由虚拟机管理程序将内存页面统一标识为脏页面(dirty page),然后通过多轮的内存拷贝过程将内存状态数据从源虚拟机迁移至目的虚拟机,迁移过程中通过位图进行控制。随着虚拟机迁移过程的执行,内存数据逐步被传送到目的虚拟机,当需要传输的内存数据量低于事先设定的门限值时停止迭代,将最后一轮产生的脏页数据全部拷贝到目的虚拟机,启动目的虚拟机,停止源虚拟机。在虚拟机迁移过程中,为了防止迭代次数过多导致时间过长,可以设置迭代次数门限值,当迭代次数超过门限值后,直接跳出循环并执行最后一轮迭代,拷贝脏页数据,启动目的虚拟机 并停止源虚拟机。
受限于网络带宽,内存拷贝可能会导致迁移时间过长,可以采用Wkdm、Delta等压缩技术对内存数据进行处理,在源虚拟机上进行内存数据压缩,在目的虚拟机上解压缩。针对迁移过程中内存数据重复冗余的特点,可利用冗余数据消除方法减少重复冗余内存数据的拷贝。
目前的开源基础设施即服务IaaS(Infrastructure as a Service)平台OpenStack支持两种类型的虚拟机迁移,即在线迁移和块迁移。在线迁移将实例信息保存在共享存储中,而块迁移需要同时迁移实例内存数据和磁盘文件数据。在虚拟机迁移过程中,OpenStack首先进行迁移条件审核,包括权限、参数、目的主机内存容量、源主机与目的虚拟机管理程序版本等。通过审核以后,调用libvirt接口,把源主机迁移到目的主机。迁移过程中,采用轮循等检测方法跟踪虚拟机迁移状态,直到迁移成功。
网络功能虚拟化环境下,虚拟网络功能运行于虚拟机VM的客户操作系统Guest OS中,当业务负载增加或缩减时,需要对虚拟网络功能进行扩容与缩容,采用的主要方法是为虚拟网络功能分配更合适的虚拟机,并将虚拟网络功能正在处理的业务数据迁移到新的虚拟机上。
但是一个具体的网络服务NS(Network Service)包括多个虚拟网络功能,业务负载的增加或减少会导致多个虚拟网络功能需要扩容或缩容,传统的方式仅仅从单个虚拟网络功能业务伸缩的角度来考虑问题,往往达不到全局最优。
发明内容
针对现有技术中存在的缺陷,本发明的目的在于提供一种能从全 局最优的角度实现虚拟网络功能的扩容与缩容的虚拟网络功能扩容与缩容方法。
为达到以上目的,本发明采取的技术方案是:
一种虚拟网络功能扩容与缩容方法,该方法包括以下步骤:
采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及
根据任意两个虚拟网络功能VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。
在上述技术方案的基础上,采集历史业务数据进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容,具体过程为:
根据采集的历史业务数据,将历史业务负载按照负载量大小分为多个级别;
在时间轴上分析历史业务负载的关联关系,分析不同时间段上历史业务负载的转移情况,获取时间序列业务关联模型;
根据时间序列业务关联模型和当前时间段的业务负载的级别,计算由当前级别的业务负载转移到其他级别的业务负载的转移概率,将最大转移概率所对应的业务负载的级别,作为与当前时间段相邻的下一个时间段的业务负载的级别;
判断当前时间段的业务负载的级别是高于还是低于下一个时间段的业务负载的级别,若高于,则进行缩容;若低于,则进行扩容。
在上述技术方案的基础上,
预先设定采用高级别优先选择算法或采用低级别优先选择算法;
当存在至少两个不同级别的业务负载的转移概率相同且最大时,若预先设定的是高级别优先选择算法,则选择其中级别高的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别;若预先设定的是低级别优先选择算法,则选择其中级别低的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别。
在上述技术方案的基础上,所述方法还包括以下步骤:从多个数据中心中选出最佳虚拟机运行扩容或缩容后的VNF。
在上述技术方案的基础上,从多个数据中心中选出最佳的虚拟机运行扩容或缩容后的VNF,具体过程为:
根据不同应用场景的需求,从多个数据中心选择最佳数据中心,所述应用场景需求包括业务负载传送时间最短、经过的VNF路径最短、安全级别最高;
根据当前所需要的VNF的类型,从最佳数据中心选择最佳服务器区域;
根据服务器空闲虚拟机的数量,从最佳服务器区域中选择空闲虚拟机的数量最多的服务器作为最佳服务器;
根据VNF所需要的虚拟核数和各个空闲虚拟机提供的核数差值,从最佳服务器中选择差值最小的空闲虚拟机作为最佳虚拟机。
在上述技术方案的基础上,根据任意两个VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,具体过程为:
A.从所有需要扩容或者缩容的VNF中任选一个VNF,加入到参考VNF组中;
B.判断剩余的每个VNF是否与参考VNF组中的某一个VNF的高维空间距离在预设的门限值内,若是,则将该VNF加入到参考 VNF组中,若否,则不进行操作;
C.按照步骤B完成对所有的剩余VNF的判断后,将最终得到的参考VNF组作为扩容组或者缩容组;
D.对于扩容组或者缩容组以外的VNF,按照步骤A至C进行处理;
E.重复步骤D,直到所有需要扩容或者缩容的VNF均处理完毕。
在上述技术方案的基础上,所述高维空间距离为切比雪夫距离。
与此同时,本发明的目的在于提供一种能从全局最优的角度实现虚拟网络功能的扩容与缩容的虚拟网络功能扩容与缩容系统。
为达到以上目的,本发明采取的技术方案是:
一种虚拟网络功能扩容与缩容系统,包括:
虚拟化编排器,其用于采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及
虚拟化功能管理器,其用于根据任意两个虚拟网络功能VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。
在上述技术方案的基础上,所述虚拟化编排器包括业务负载管理器,所述业务负载管理器采集历史业务数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容,具体过程为:
根据采集的历史业务数据,将历史业务负载按照负载量大小分为多个级别;
在时间轴上分析历史业务负载的关联关系,分析不同时间段上历史业务负载的转移情况,获取时间序列业务关联模型;
根据时间序列业务关联模型和当前时间段的业务负载的级别,计算由当前级别的业务负载转移到其他级别的业务负载的转移概率,将最大转移概率所对应的业务负载的级别作为与当前时间段相邻的下一个时间段的业务负载的级别;
判断当前时间段的业务负载的级别是高于还是低于下一个时间段的业务负载的级别,若高于,则进行缩容;若低于,则进行扩容。
在上述技术方案的基础上,
所述业务负载管理器预先设定采用高级别优先选择算法或采用低级别优先选择算法;
当存在至少两个不同级别的业务负载的转移概率相同且最大时,若预先设定的是高级别优先选择算法,则选择其中级别高的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别;若预先设定的是低级别优先选择算法,则选择其中级别低的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别。
在上述技术方案的基础上,所述虚拟化编排器还用于从多个数据中心中选出最佳虚拟机运行扩容或缩容后的VNF。
在上述技术方案的基础上,所述虚拟化编排器从多个数据中心中选出最佳的虚拟机运行扩容或缩容后的VNF,具体过程为:
根据不同应用场景的需求,从多个数据中心选择最佳数据中心,所述应用场景需求包括业务负载传送时间最短、经过的VNF路径最短、安全级别最高;
根据当前所需要的VNF的类型,从最佳数据中心选择最佳服务器区域;
根据服务器空闲虚拟机的数量,从最佳服务器区域中选择空闲虚拟机的数量最多的服务器作为最佳服务器;
根据VNF所需要的虚拟核数和各个空闲虚拟机提供的核数差值,从最佳服务器中选择差值最小的空闲虚拟机作为最佳虚拟机。
在上述技术方案的基础上,所述虚拟化功能管理器根据任意两个VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,具体过程为:
A.从所有需要扩容或者缩容的VNF中任选一个VNF,加入到参考VNF组中;
B.判断剩余的每个VNF是否与参考VNF组中的某一个VNF的高维空间距离在预设的门限值内,若是,则将该VNF加入到参考VNF组中,若否,则不进行操作;
C.按照步骤B完成对所有的剩余VNF的判断后,将最终得到的参考VNF组作为扩容组或者缩容组;
D.对于扩容组或者缩容组以外的VNF,按照步骤A至C进行处理;
E.重复步骤D,直到所有需要扩容或者缩容的VNF均处理完毕。
在上述技术方案的基础上,所述高维空间距离为切比雪夫距离。
与现有技术相比,本发明的优点在于:
相对于传统的方式仅仅从单个虚拟网络功能业务伸缩的角度来考虑问题,往往达不到全局最优。本发明的虚拟网络功能扩容与缩容方法提出动态组合方法实现虚拟网络功能迁移,面向网络服务,对虚拟网络功能的关联关系进行分析,得到最优的扩容与缩容组合集,然后再对同一扩容组中的虚拟网络功能同时进行扩容或缩容,从而实现 了从全局最优的角度实现虚拟网络功能的扩容与缩容。
图1为本发明实施例中虚拟网络功能扩容与缩容方法的流程图;
图2为本发明实施例中虚拟网络功能扩容与缩容方法中步骤S1的流程图;
图3为本发明实施例中虚拟网络功能扩容与缩容系统框图;
图4为本发明实施例中最佳虚拟机选择示意图;
图5为本发明实施例中虚拟网络功能动态组合式扩容或缩容示意图。
以下结合附图及实施例对本发明作进一步详细说明。
实施例1:
参见图1所示,本实施例提供一种虚拟网络功能扩容与缩容方法,该方法包括以下步骤:
S1.采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及
S2.根据任意两个虚拟网络功能VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。
相对于传统的方式,其仅仅从单个虚拟网络功能业务伸缩的角度来考虑问题,往往达不到全局最优。本实施例提出动态组合方法实现虚拟网络功能迁移,面向网络服务,对虚拟网络功能的关联关系进行分析,得到最优的扩容与缩容组合集,然后再对同一扩容组中的虚拟 网络功能同时进行扩容或缩容,从而实现了从全局最优的角度实现虚拟网络功能的扩容与缩容。
实施例2:
参见图2所示,作为一个较好的可选方式,在实施例1的基础上,采集历史业务数据进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容,具体过程为:
S11.根据采集的历史业务数据,将历史业务负载按照负载量大小分为多个级别;
优选的,本实施例将业务负载分为超高负载UH(Ultra High Level)、高负载H(High Level)、普通负载C(Common Level)、低负载L(Low Level)、超低负载UL(Ultra Low Level)五个级别。在具体实践中,可以根据业务负载数量调整分类数目,例如在负载量变化范围较大时划分为七级或者九级,在负载量变化范围较小时调整为三级。
S12.在时间轴上分析历史业务负载的关联关系,分析不同时间段上历史业务负载的转移情况,获取时间序列业务关联模型;
根据业务负载预测模型,业务负载管理器建立一个未来业务负载分类序列。本发明采用符号
表示这个分类序列,其中上标R表示序列范围Range,下标I表示虚拟网络功能标识Identifier,例如虚拟网络功能VNF3的未来时间段的五个分类负载序列可以表示为
也即是说VNF3业务负载趋势为高负载、超高负载、超高负载、高负载、超高负载。然后在时间轴上分析业务负载的关联关系,分析各个不同的时间片段上业务负载的相似性,包括高峰段、节假日段、突发流量段等,得到量化后的时间序列业务负载关联模型。基于历史数据的业务负载关联模型统计各个时 间序列点上的概率值,得到业务负载转移模型。
S13.根据时间序列业务关联模型和当前时间段的业务负载的级别,计算由当前级别的业务负载转移到其他级别的业务负载的转移概率,将最大转移概率所对应的业务负载的级别,作为与当前时间段相邻的下一个时间段的业务负载的级别;
进一步地,预先设定采用高级别优先选择算法或采用低级别优先选择算法;
当存在至少两个不同级别的业务负载的转移概率相同且最大时,若预先设定的是高级别优先选择算法,则选择其中级别高的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别;若预先设定的是低级别优先选择算法,则选择其中级别低的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别。
具体而言,根据上述得到的
通过分析得知从高负载有可能会转移到超高负载UH和普通负载C,转移序列为H→UH,H→C,H→UH,H→UH。计算概率得
这两个概率表示当前处于高负载H的情况下,下一步有75%的可能转移至超高负载UH,有25%的可能性转移至普通负载C。
计算得到业务负载转移序列后,通过业务负载选择算法得到下一步的可能业务负载量。如果转移概率值可以进行比较,则选择较大业务负载转移概率值作为下一步的业务负载量。例如,对于
和
这两个转移概率值,本发明将选取概率值0.75对应的负载量,即超高负载作为下一步的业务负载量。如果出现两个转移概率值 相等的情况,如
和
本发明采用两种选择策略。业务负载高级别优先选择算法将返回
对应的业务负载,即超高业务负载,而业务负载低级别优先选择算法将返回
对应的业务负载,即普通业务负载。
S14.判断当前时间段的业务负载的级别是高于还是低于下一个时间段的业务负载的级别,若高于,则进行缩容;若低于,则进行扩容。
实施例3:
作为一个较好的可选方式,在实施例1的基础上,所述方法还包括以下步骤:从多个数据中心中选出最佳虚拟机运行扩容或缩容后的VNF。其中数据中心指的是运营商云化网络数据中心,运营商网络云化转型以后,面向云化网络构建的数据中心,部署电信云平台、虚拟网络功能、管理编排平台等,在本发明中简称数据中心。
进一步地,参见图4所示,从多个数据中心中选出最佳的虚拟机运行扩容或缩容后的VNF,具体过程为:
根据不同应用场景的需求,从多个数据中心选择最佳数据中心,所述应用场景需求包括业务负载传送时间最短、经过的VNF路径最短、安全级别最高;
具体而言,假设当前有三个数据中心DC
1、DC
2、DC
3(DC:Data Center数据中心),基于不同应用场景的需求,选择函数f
DC会采用不同的策略,例如业务负载传送时间最短,经过的虚拟网络功能最短,安全级别最高等,然后选择出最优的数据中心。本实施例中的选择函数f
DC为通用选择函数,不涉及具体的特定的数据中心选择逻辑。假设当前三个数据中心,需要扩容的虚拟网络功能处于数据中心DC
1,业务负载会由高负载转移至超高负载,但数据中心DC
1没有对应超高 业务负载量的虚拟机,因此需要从数据中心DC
2和DC
3中选择,假设数据中心DC
2比数据中心DC
3的业务负载传送时间要短,则本次选择结果为DC
2,即DC
2=f
DC(DC
1,DC
2,DC
3)。
根据当前所需要的VNF的类型,从最佳数据中心选择最佳服务器区域;
具体而言,选择出最佳数据中心后,本发明通过区域选择函数f
RG来选择合适的服务器区域。数据中心不同区域的服务器适合不同类型的虚拟网络功能,假设区域一RG
1(RG:Region区域)适合承载高带宽视频类业务的虚拟网络功能,区域二RG
2适合承载分布式并行海量计算类业务的虚拟网络功能,区域三RG
3适合承载低延迟远程医疗业务的虚拟网络功能,区域四RG
4适合承载安全防护业务的虚拟网络功能,如果用户需要虚拟网络功能承载海量数据计算业务,则区域选择函数返回结果为区域三,即RG
3=f
RG(RG
1,RG
2,RG
3,RG
4)。
根据服务器空闲虚拟机的数量,从最佳服务器区域中选择空闲虚拟机的数量最多的服务器作为最佳服务器;
具体而言,每一个区域包括多台服务器,服务器选择函数f
SRV返回合适的服务器支撑智能化扩容或缩容。本发明采用空闲虚拟机最多选择算法,即函数f
SRV返回空闲虚拟机数量最多的服务器。如果出现空闲虚拟机数量最多的服务器有多台,则随机选择一台服务器。假如四台服务器SRV
1,SRV
2,SRV
3,SRV
4(SRV:Server服务器)的空闲虚拟机数量分别为8、6、3,12,则服务器选择函数返回服务器四SRV
4,即SRV
4=f
SRV(SRV
1,SRV
2,SRV
3,SRV
4)。
根据VNF所需要的虚拟核数和各个空闲虚拟机提供的核数差值,从最佳服务器中选择差值最小的空闲虚拟机作为最佳虚拟机。
具体而言,本实施例通过虚拟机选择函数f
VM来选择合适的虚拟 机,f
VM采用满足需求的最小虚拟核数适配方法,即虚拟网络功能所需要的虚拟核数和各个空闲虚拟机提供的核数差值,返回差值最小的那台空闲虚拟机,这种选择方法目的是保留虚拟核数多的虚拟机用于其他应用。假如虚拟网络功能需要8个虚拟核,当前服务器三个虚拟机VM
1,VM
2,VM
3(VM:Virtual Machine虚拟机)的虚拟核数分别为12,10,6,虚拟机选择函数f
VM首先进行判断,发现虚拟机VM
3只能提供6个虚拟核,达不到虚拟网络功能需求,排除虚拟机VM
3。然后计算虚拟机VM
1和虚拟机VM
2能够提供的虚拟核心与虚拟网络功能需要的核数之差,发现虚拟机VM
2的差值最小,因此返回虚拟机VM
2,即VM
2=f
VM(VM1,VM2,VM3)。
综上所述,本实施例将通过四个函数f
DC、f
RG、f
SRV、f
VM分别选择数据中心、数据中心的特定区域、区域中的特定服务器、服务器上的特定虚拟机。其选择过程分别为DC
2=f
DC(DC
1,DC
2,DC
3),RG
3=f
RG(RG
1,RG
2,RG
3,RG
4),SRV
4=f
SRV(SRV
1,SRV
2,SRV
3,SRV
4),VM
2=f
VM(VM
1,VM
2,VM
3),最后选择虚拟机VM
2运行扩容后的虚拟网络功能,整个过程描述为VM
2=f
VM。f
SRV。f
RG。f
DC(DC
1,DC
2,DC
3)。
实施例4:
作为一个较好的可选方式,在实施例1的基础上,根据任意两个VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,具体过程为:
A.从所有需要扩容或者缩容的VNF中任选一个VNF,加入到参考VNF组中;
B.判断剩余的每个VNF是否与参考VNF组中的某一个VNF 的高维空间距离在预设的门限值内,若是,则将该VNF加入到参考VNF组中,若否,则不进行操作;
C.按照步骤B完成对所有的剩余VNF的判断后,将最终得到的参考VNF组作为扩容组或者缩容组;
D.对于扩容组或者缩容组以外的VNF,按照步骤A至C进行处理;
E.重复步骤D,直到所有需要扩容或者缩容的VNF均处理完毕。
高维空间距离可以采用曼哈顿距离、欧式距离、明可夫斯基距离、切比雪夫距离等方式计算,优选切比雪夫距离。
具体而言,参见图5所示,图5描述了虚拟网络功能动态组合式扩容和缩容。图5最下部分为服务器,通过虚拟化引擎构建计算、存储、网络三类虚拟化资源池,在虚拟机上运行虚拟网络功能VNF。本实施例提出基于切比雪夫距离的扩缩容组选择算法,基于扩缩容时间、虚拟连接VL带宽计算切比雪夫距离(为了提高精度,避免不同参数的量纲差异导致数据不准,可以采用归一化、或调整量纲后的参数值计算。),将切比雪夫距离与虚拟化编排器NFVO设定的门限值对比,小于门限值的虚拟网络功能为一个扩缩容组。门限值由虚拟化编排器NFVO根据计算得到。本实施例采用二维数据(tim,ban)表示虚拟网络功能VNF扩缩容时间、本虚拟网络功能关联的所有VL的最大带宽。假如图5中的四个虚拟网络功能VNF-1、VNF-2、VNF-3、VNF-4的扩缩容时间和关联VL最大带宽分别表示为(tim1,ban1),(tim2,ban2),(tim3,ban3),(tim4,ban4),分别计算四个虚拟网络功能两两之间的六个切比雪夫距离,例如计算虚拟网络功能VNF-1与虚拟网络功能VNF-2之间切比雪夫距离d(1,2)=max(|tim1-tim2|,|ban1-ban2|),同样计算出其他五个距离d(1,3),d(1,4),d(2, 3),d(2,4),d(3,4)。计算出六个距离以后,首先将虚拟网络功能VNF-1选进扩缩容组,然后比较d(1,2)与门限值,如果d(1,2)大于门限值,则不会将VNF-2选进扩缩容组。接着比较d(1,3)与门限值,如果d(1,3)小于门限值,则将VNF-3选进扩缩容组。此时近缩容组包括两个虚拟网络功能VNF-1和VNF-3。接着判断是否将虚拟网络功能VNF-4选进扩缩容组内。分别查看VNF-4与扩缩容组内VNF-1、VNF-3的切比雪夫距离是否小于门限值,即分别比较d(1,4)和d(3,4)与门限值的大小,只要有一个距离小于门限值,就将虚拟网络功能添加到扩缩容组中。得到扩容与缩容组后,本实施例同时将扩缩容组中的虚拟网络功能进行扩容与缩容。在图5中,网络功能虚拟化编排器NFVO与虚拟网络功能管理器VNFM启动扩容过程,同时对VNF-1、VNF-3、VNF-4进行扩容,并对整个扩容过程进行管理。
实施例5:
参见图3所示,本实施例提供一种虚拟网络功能扩容与缩容系统,包括:
虚拟化编排器,其用于采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及
虚拟化功能管理器,其用于根据任意两个虚拟网络功能VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。
具体而言,业务负载管理器位于网络功能虚拟化编排器NFVO (Network Function Virtualization Orchestrator)中,与各个虚拟网络功能管理器VNFM(Virtualized Network Function Manager)相连,获取各个虚拟网络功能VNF(Virtualized Network Function)中的业务负载数据量。本实施例中,虚拟网络功能管理器VNFM上传的只是业务负载量化指标,并不需要对负载数据进行镜像和上传。业务数据采集周期由管理员在网络功能虚拟化编排器NFVO中设定并下发至VNFM和VNF中。本实施例中,网络功能虚拟化环境包括一个统一的网络功能虚拟化编排器NFVO,然后对应多个数据中心,包括网络功能虚拟化基础设施NFVI(Network Function Virtualization Infrastructure)和虚拟化基础设施管理器VIM(Virtualized Infrastructure Manager)。每个数据中心运行一套或多套虚拟网络功能组VNFGrp,每个虚拟网络功能组VNFGrp包括一个虚拟网络功能管理器VNFM,一个虚拟网元管理器EM(Element Manger),还有多个虚拟网络功能VNF。在具体实践中,可以基于同一个厂商的业务功能实体构建一个虚拟网络功能组VNFGrp。
参见图3所示,其描述了一个业务负载管理器连接两个VNFM(即VNFMA与VNFMB)的例子,其中VNFMA管理虚拟网络功能VNF-A1和VNF-A2,VNFMB管理虚拟网络功能VNF-B1与VNF-B2。本实施例中,虚拟网络功能组VNFGrpA包括四个实体,分别为虚拟网络功能管理器VNFMA、虚拟网元管理EMA、虚拟网络功能VNF-A1、虚拟网络功能VNF-A2。在具体实施过程中,同一厂商的虚拟化功能实体可以构成一个虚拟网络功能组。图5下部为网络功能虚拟化基础设施NFVI和虚拟基础设施管理器VIM,分别实现硬件虚拟化和虚拟计算、存储、网络资源的管理。
业务负载流量从图3中的四个虚拟网络功能(即VNF-A1、 VNF-A2、VNF-B1、VNF-B2)产生,负载流量按设定的周期上传虚拟网络功能管理器VNFM,并上传至业务负载管理器。在实际实施过程中,业务负载管理器会根据具体的应用场景,分析历史业务负载数据,确定合适的业务负载采集周期并下发至各个虚拟网络功能实体。虚拟网络功能所消耗的资源情况由虚拟基础设施管理器VIM采集并上报网络功能虚拟化编排器NFVO,然后交给业务负载管理器进行分析。
实施例6:
作为一个较好的可选方式,在实施例5的基础上,所述业务负载管理器采集历史业务数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容,具体过程为:
根据采集的历史业务数据,将历史业务负载按照负载量大小分为多个级别;
在时间轴上分析历史业务负载的关联关系,分析不同时间段上历史业务负载的转移情况,获取时间序列业务关联模型;
根据时间序列业务关联模型和当前时间段的业务负载的级别,计算由当前级别的业务负载转移到其他级别的业务负载的转移概率,将最大转移概率所对应的业务负载的级别作为与当前时间段相邻的下一个时间段的业务负载的级别;
进一步地,所述业务负载管理器预先设定采用高级别优先选择算法或采用低级别优先选择算法;
当存在至少两个不同级别的业务负载的转移概率最大时,若预先设定的是高级别优先选择算法,则选择其中级别高的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别;若预先设定的是低级别优先选择算法,则选择其中级别低的业务负载作为与当前时间 段相邻的下一个时间段的业务负载的级别。
判断当前时间段的业务负载的级别是高于还是低于下一个时间段的业务负载的级别,若高于,则进行缩容;若低于,则进行扩容。
实施例7:
作为一个较好的可选方式,在实施例5的基础上,所述虚拟化编排器还用于从多个数据中心中选出最佳虚拟机运行扩容或缩容后的虚拟网络功能VNF。
进一步地,所述虚拟化编排器从多个数据中心中选出最佳的虚拟机运行扩容或缩容后的VNF,具体过程为:
根据不同应用场景的需求,从多个数据中心选择最佳数据中心,所述应用场景需求包括业务负载传送时间最短、经过的VNF路径最短、安全级别最高;
根据当前所需要的VNF的类型,从最佳数据中心选择最佳服务器区域;
根据服务器空闲虚拟机的数量,从最佳服务器区域中选择空闲虚拟机的数量最多的服务器作为最佳服务器;
根据VNF所需要的虚拟核数和各个空闲虚拟机提供的核数差值,从最佳服务器中选择差值最小的空闲虚拟机作为最佳虚拟机。
实施例8:
作为一个较好的可选方式,在实施例5的基础上,所述虚拟化功能管理器根据任意两个VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,具体过程为:
A.从所有需要扩容或者缩容的VNF中任选一个VNF,加入到 参考VNF组中;
B.判断剩余的每个VNF是否与参考VNF组中的某一个VNF的高维空间距离在预设的门限值内,若是,则将该VNF加入到参考VNF组中,若否,则不进行操作;
C.按照步骤B完成对所有的剩余VNF的判断后,将最终得到的参考VNF组作为扩容组或者缩容组;
D.对于扩容组或者缩容组以外的VNF,按照步骤A至C进行处理;
E.重复步骤D,直到所有需要扩容或者缩容的VNF均处理完毕。
高维空间距离可以采用曼哈顿距离、欧式距离、明可夫斯基距离、切比雪夫距离等方式计算,本实施例优选切比雪夫距离。
本发明不局限于上述实施方式,对于本技术领域的普通技术人员来说,在不脱离本发明原理的前提下,还可以做出若干改进和润饰,这些改进和润饰也视为本发明的保护范围之内。本说明书中未作详细描述的内容属于本领域专业技术人员公知的现有技术。
Claims (14)
- 一种虚拟网络功能扩容与缩容方法,其特征在于,该方法包括以下步骤:采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及根据任意两个虚拟网络功能VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。
- 如权利要求1所述的虚拟网络功能扩容与缩容方法,其特征在于,采集历史业务数据进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容,具体过程为:根据采集的历史业务数据,将历史业务负载按照负载量大小分为多个级别;在时间轴上分析历史业务负载的关联关系,分析不同时间段上历史业务负载的转移情况,获取时间序列业务关联模型;根据时间序列业务关联模型和当前时间段的业务负载的级别,计算由当前级别的业务负载转移到其他级别的业务负载的转移概率,将最大转移概率所对应的业务负载的级别,作为与当前时间段相邻的下一个时间段的业务负载的级别;判断当前时间段的业务负载的级别是高于还是低于下一个时间段的业务负载的级别,若高于,则进行缩容;若低于,则进行扩容。
- 如权利要求2所述的虚拟网络功能扩容与缩容方法,其特征在于:预先设定采用高级别优先选择算法或采用低级别优先选择算法;当存在至少两个不同级别的业务负载的转移概率相同且最大时,若预先设定的是高级别优先选择算法,则选择其中级别高的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别;若预先设定的是低级别优先选择算法,则选择其中级别低的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别。
- 如权利要求1所述的虚拟网络功能扩容与缩容方法,其特征在于,所述方法还包括以下步骤:从多个数据中心中选出最佳虚拟机运行扩容或缩容后的VNF。
- 如权利要求4所述的虚拟网络功能扩容与缩容方法,其特征在于,从多个数据中心中选出最佳的虚拟机运行扩容或缩容后的VNF,具体过程为:根据不同应用场景的需求,从多个数据中心选择最佳数据中心,所述应用场景需求包括业务负载传送时间最短、经过的VNF路径最短、安全级别最高;根据当前所需要的VNF的类型,从最佳数据中心选择最佳服务器区域;根据服务器空闲虚拟机的数量,从最佳服务器区域中选择空闲虚拟机的数量最多的服务器作为最佳服务器;根据VNF所需要的虚拟核数和各个空闲虚拟机提供的核数差值,从最佳服务器中选择差值最小的空闲虚拟机作为最佳虚拟机。
- 如权利要求1所述的虚拟网络功能扩容与缩容方法,其特征在于,根据任意两个VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,具体过程为:A.从所有需要扩容或者缩容的VNF中任选一个VNF,加入到 参考VNF组中;B.判断剩余的每个VNF是否与参考VNF组中的某一个VNF的高维空间距离在预设的门限值内,若是,则将该VNF加入到参考VNF组中,若否,则不进行操作;C.按照步骤B完成对所有的剩余VNF的判断后,将最终得到的参考VNF组作为扩容组或者缩容组;D.对于扩容组或者缩容组以外的VNF,按照步骤A至C进行处理;E.重复步骤D,直到所有需要扩容或者缩容的VNF均处理完毕。
- 如权利要求6所述的虚拟网络功能扩容与缩容方法,其特征在于:所述高维空间距离为切比雪夫距离。
- 一种虚拟网络功能扩容与缩容系统,其特征在于,包括:虚拟化编排器,其用于采集历史业务负载数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容;以及虚拟化功能管理器,其用于根据任意两个虚拟网络功能VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,对每个扩容组中包含的VNF同时进行扩容,或对每个缩容组中包含的VNF同时进行缩容。
- 如权利要求8所述的虚拟网络功能扩容与缩容系统,其特征在于,所述虚拟化编排器包括业务负载管理器,所述业务负载管理器采集历史业务数据并进行分析,对未来时间段的业务负载进行预测,根据预测结果判断是进行扩容还是缩容,具体过程为:根据采集的历史业务数据,将历史业务负载按照负载量大小分为 多个级别;在时间轴上分析历史业务负载的关联关系,分析不同时间段上历史业务负载的转移情况,获取时间序列业务关联模型;根据时间序列业务关联模型和当前时间段的业务负载的级别,计算由当前级别的业务负载转移到其他级别的业务负载的转移概率,将最大转移概率所对应的业务负载的级别作为与当前时间段相邻的下一个时间段的业务负载的级别;判断当前时间段的业务负载的级别是高于还是低于下一个时间段的业务负载的级别,若高于,则进行缩容;若低于,则进行扩容。
- 如权利要求9所述的虚拟网络功能扩容与缩容系统,其特征在于:所述业务负载管理器预先设定采用高级别优先选择算法或采用低级别优先选择算法;当存在至少两个不同级别的业务负载的转移概率相同且最大时,若预先设定的是高级别优先选择算法,则选择其中级别高的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别;若预先设定的是低级别优先选择算法,则选择其中级别低的业务负载作为与当前时间段相邻的下一个时间段的业务负载的级别。
- 如权利要求8所述的虚拟网络功能扩容与缩容系统,其特征在于,所述虚拟化编排器还用于从多个数据中心中选出最佳虚拟机运行扩容或缩容后的VNF。
- 如权利要求11所述的虚拟网络功能扩容与缩容系统,其特征在于,所述虚拟化编排器从多个数据中心中选出最佳的虚拟机运行扩容或缩容后的VNF,具体过程为:根据不同应用场景的需求,从多个数据中心选择最佳数据中心, 所述应用场景需求包括业务负载传送时间最短、经过的VNF路径最短、安全级别最高;根据当前所需要的VNF的类型,从最佳数据中心选择最佳服务器区域;根据服务器空闲虚拟机的数量,从最佳服务器区域中选择空闲虚拟机的数量最多的服务器作为最佳服务器;根据VNF所需要的虚拟核数和各个空闲虚拟机提供的核数差值,从最佳服务器中选择差值最小的空闲虚拟机作为最佳虚拟机。
- 如权利要求8所述的虚拟网络功能扩容与缩容系统,其特征在于,所述虚拟化功能管理器根据任意两个VNF之间的高维空间距离,确定VNF相互之间的关联程度,将需要扩容或缩容的多个VNF按照关联程度进行分组,形成多个扩容组或缩容组,具体过程为:A.从所有需要扩容或者缩容的VNF中任选一个VNF,加入到参考VNF组中;B.判断剩余的每个VNF是否与参考VNF组中的某一个VNF的高维空间距离在预设的门限值内,若是,则将该VNF加入到参考VNF组中,若否,则不进行操作;C.按照步骤B完成对所有的剩余VNF的判断后,将最终得到的参考VNF组作为扩容组或者缩容组;D.对于扩容组或者缩容组以外的VNF,按照步骤A至C进行处理;E.重复步骤D,直到所有需要扩容或者缩容的VNF均处理完毕。
- 如权利要求13所述的虚拟网络功能扩容与缩容系统,其特征在于:所述高维空间距离为切比雪夫距离。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810943325.1 | 2018-08-17 | ||
| CN201810943325.1A CN109189552B (zh) | 2018-08-17 | 2018-08-17 | 虚拟网络功能扩容与缩容方法及系统 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2020034492A1 true WO2020034492A1 (zh) | 2020-02-20 |
Family
ID=64918322
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2018/119286 Ceased WO2020034492A1 (zh) | 2018-08-17 | 2018-12-05 | 虚拟网络功能扩容与缩容方法及系统 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109189552B (zh) |
| WO (1) | WO2020034492A1 (zh) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111093099B (zh) * | 2019-02-15 | 2023-04-18 | 杭州海康威视系统技术有限公司 | 一种流媒体服务调度方法、装置及系统 |
| CN112751685B (zh) * | 2019-10-29 | 2023-04-07 | 中国电信股份有限公司 | 网络切片用户面容量配置方法、装置和系统 |
| CN112732409B (zh) * | 2021-01-21 | 2022-07-22 | 上海交通大学 | 一种vnf架构下使能零耗时网络流负载均衡的方法及装置 |
| CN113014436B (zh) * | 2021-03-15 | 2022-10-11 | 中国联合网络通信集团有限公司 | 网络服务管理方法、装置和计算机存储介质 |
| CN113342561A (zh) * | 2021-06-18 | 2021-09-03 | 上海哔哩哔哩科技有限公司 | 任务诊断方法及系统 |
| CN113918341B (zh) * | 2021-10-20 | 2024-10-25 | 中国平安人寿保险股份有限公司 | 设备调度方法、装置、设备及存储介质 |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104469798A (zh) * | 2014-12-12 | 2015-03-25 | 重庆邮电大学 | 一种基于马尔可夫链的通信网络负载状态信息预测方法 |
| CN105610592A (zh) * | 2014-10-30 | 2016-05-25 | 中兴通讯股份有限公司 | 一种部署虚拟核心网的方法和装置 |
| US20160366014A1 (en) * | 2015-06-09 | 2016-12-15 | Kt Corporation | Method and apparatus for network function virtualization |
| CN106301829A (zh) * | 2015-05-21 | 2017-01-04 | 华为技术有限公司 | 一种网络业务扩容的方法和装置 |
| US20170093721A1 (en) * | 2015-09-25 | 2017-03-30 | Fujitsu Limited | Setting method, server device and service chain system |
| CN108243110A (zh) * | 2016-12-26 | 2018-07-03 | 华为技术有限公司 | 一种资源调整方法、装置和系统 |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104009467B (zh) * | 2014-05-13 | 2016-07-06 | 南京邮电大学 | 一种计及预安排停电影响的配电网可靠性评估和预测方法 |
| CN106161076B (zh) * | 2015-04-22 | 2019-06-11 | 华为技术有限公司 | 虚拟网络功能扩展方法和装置 |
| US9992609B2 (en) * | 2015-06-01 | 2018-06-05 | Huawei Technologies Co., Ltd. | Method and system for MTC event management |
| CN106452818B (zh) * | 2015-08-13 | 2020-01-21 | 阿里巴巴集团控股有限公司 | 一种资源调度的方法和系统 |
| CN106484540B (zh) * | 2016-10-20 | 2019-11-19 | 腾讯科技(深圳)有限公司 | 一种资源配置方法及装置 |
| CN106775936B (zh) * | 2016-11-30 | 2020-06-26 | 华为技术有限公司 | 一种虚拟机的管理方法及装置 |
| CN107145393A (zh) * | 2017-04-27 | 2017-09-08 | 努比亚技术有限公司 | 一种负载调整方法、设备及计算机可读存储介质 |
-
2018
- 2018-08-17 CN CN201810943325.1A patent/CN109189552B/zh active Active
- 2018-12-05 WO PCT/CN2018/119286 patent/WO2020034492A1/zh not_active Ceased
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105610592A (zh) * | 2014-10-30 | 2016-05-25 | 中兴通讯股份有限公司 | 一种部署虚拟核心网的方法和装置 |
| CN104469798A (zh) * | 2014-12-12 | 2015-03-25 | 重庆邮电大学 | 一种基于马尔可夫链的通信网络负载状态信息预测方法 |
| CN106301829A (zh) * | 2015-05-21 | 2017-01-04 | 华为技术有限公司 | 一种网络业务扩容的方法和装置 |
| US20160366014A1 (en) * | 2015-06-09 | 2016-12-15 | Kt Corporation | Method and apparatus for network function virtualization |
| US20170093721A1 (en) * | 2015-09-25 | 2017-03-30 | Fujitsu Limited | Setting method, server device and service chain system |
| CN108243110A (zh) * | 2016-12-26 | 2018-07-03 | 华为技术有限公司 | 一种资源调整方法、装置和系统 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109189552A (zh) | 2019-01-11 |
| CN109189552B (zh) | 2020-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020034492A1 (zh) | 虚拟网络功能扩容与缩容方法及系统 | |
| CN110134495B (zh) | 一种容器跨主机在线迁移方法、存储介质及终端设备 | |
| CN111381928B (zh) | 一种虚拟机迁移方法、云计算管理平台和存储介质 | |
| US12373405B2 (en) | Data migration method and apparatus, device, medium, and computer product | |
| US9954758B2 (en) | Virtual network function resource allocation and management system | |
| CN102694868B (zh) | 一种集群系统实现及任务动态分配方法 | |
| CN107404523A (zh) | 云平台自适应资源调度系统和方法 | |
| CN102713893A (zh) | 用于企业数据中心的虚拟化和整合分析引擎 | |
| Wang et al. | Hybrid pulling/pushing for i/o-efficient distributed and iterative graph computing | |
| CN117724853A (zh) | 基于人工智能的数据处理方法及装置 | |
| CN118484277A (zh) | 任务调度方法、任务调度系统及计算机存储介质 | |
| CN114741160B (zh) | 一种基于平衡能耗与服务质量的动态虚拟机整合方法和系统 | |
| CN120569729A (zh) | 自平衡混合专家 | |
| Karthikeyan et al. | Saldeft: Self-adaptive learning differential evolution based optimal physical machine selection for fault tolerance problem in cloud | |
| CN105635285B (zh) | 一种基于状态感知的vm迁移调度方法 | |
| US10594620B1 (en) | Bit vector analysis for resource placement in a distributed system | |
| Yang et al. | Automatic and Scalable Data Replication Manager in Distributed Computation and Storage Infrastructure of Cyber-Physical Systems. | |
| CN115567387A (zh) | 一种多维资源均衡的虚拟网络功能部署方法及系统 | |
| CN117880288A (zh) | 数据均衡方法及相关设备 | |
| Adgaonkar et al. | Availability-Aware Multi-Objective Cluster Allocation Optimization in Energy-Efficient Datacenters | |
| CN104683480A (zh) | 一种基于应用的分布式计算方法 | |
| Zhiyong et al. | An improved container cloud resource scheduling strategy | |
| Esnault et al. | Energy-aware distributed ant colony based virtual machine consolidation in IaaS Clouds bibliographic study | |
| CN118152114A (zh) | 一种煤矿地学大数据处理系统及方法 | |
| Malhotra | A critical survey of virtual machine migration techniques in cloud computing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 18930393 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 18930393 Country of ref document: EP Kind code of ref document: A1 |