WO2016197301A1 - Nfv系统中的策略协调方法和装置 - Google Patents

Nfv系统中的策略协调方法和装置 Download PDF

Info

Publication number
WO2016197301A1
WO2016197301A1 PCT/CN2015/080972 CN2015080972W WO2016197301A1 WO 2016197301 A1 WO2016197301 A1 WO 2016197301A1 CN 2015080972 W CN2015080972 W CN 2015080972W WO 2016197301 A1 WO2016197301 A1 WO 2016197301A1
Authority
WO
WIPO (PCT)
Prior art keywords
policy
policies
priority
coordination
dimension
Prior art date
Application number
PCT/CN2015/080972
Other languages
English (en)
French (fr)
Inventor
夏海涛
余芳
朱雷
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to PCT/CN2015/080972 priority Critical patent/WO2016197301A1/zh
Priority to CN201580042760.1A priority patent/CN106576260B/zh
Publication of WO2016197301A1 publication Critical patent/WO2016197301A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition

Definitions

  • the embodiments of the present invention relate to communication technologies, and in particular, to a method and device for coordinating policies in a Network Function Virtualization (NFV) system.
  • NFV Network Function Virtualization
  • NFV uses a variety of general-purpose hardware such as x86 and virtualization technology to carry software processing for many functions, thereby reducing the cost of expensive dedicated equipment deployment in the network.
  • NFV mainly has the following three characteristics: 1.
  • the software that defines the network function is completely separated from the general hardware server, memory and network switch. 2.
  • Automated orchestration Fully automated remote installation and management of software functions based on common hardware.
  • the Management and Orchestration (MANO) function of NFV can apply policy-based management.
  • NFV Orchestrator configures the automatic expansion/reduction strategy (Auto-scale-policy) of the network service instance (NS Instance).
  • the automatic capacity expansion/reduction capacity policy indicates that the NFVO automatically performs the expansion/contraction operation of the NS under specific conditions.
  • NFVO can also configure the automatic expansion/reduction strategy of the VNF instance to the VNF Manager (VNFM) through the Nfvo-Vnfm interface. , instructs the VNFM to automatically perform the VNF expansion/contraction operation under certain conditions, and increase or decrease the virtual resources of the VNF.
  • VNFM VNF Manager
  • NFVO or VNFM lifecycle management strategies
  • self-healing strategy for self-healing function configuration of NS instance load balancing strategy for VNF instance, and so on.
  • Multiple policies are generated under different conditions. When multiple policies are executed in the same functional entity, potential execution conflicts are formed. Execution conflicts between different policies may cause inconsistencies with policy makers (for example, NFV operators). Execution effect.
  • the embodiments of the present invention provide a method and a device for coordinating policies in an NFV system, which can prevent an execution conflict of a policy from causing an inconsistent execution result with a policy maker.
  • a first aspect of the present invention provides a policy coordination method in an NFV system, including:
  • the policy enforcement entity acquires M policies, and priorities of each of the M policies in a policy coordination dimension, the policy includes: a policy condition and an execution action, the mth of the M policies
  • the policy coordination dimension of the policy includes at least one of the following information: the mth policy, an execution action of the mth policy, or an optimization target of the mth policy, and the optimization goal of the mth policy is
  • the optimization target in the policy condition of the mth policy, M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M;
  • the policy execution entity is in the policy coordination dimension according to the first policy
  • the priority and the priority of the second policy under the policy coordination dimension determine a target execution action and perform the target execution action.
  • the policy enforcement entity is based on the priority of the first policy in the policy coordination dimension and the second policy in the policy
  • the priority under the coordination dimension determines the target execution action and performs the target execution action, including:
  • the policy enforcement entity compares a priority of the first policy in the policy coordination dimension and a priority of the second policy in the policy coordination dimension;
  • the policy enforcement entity determines that the execution action of the first policy is The target performs an action and performs an execution action of the first policy
  • the policy enforcement entity determines that the execution action of the second policy is The target performs an action and performs an execution action of the second policy.
  • the policy enforcement entity acquires M policies, and priorities of the policy coordination dimensions of each of the M policies, including:
  • the policy enforcement entity receives the M policies sent by the policy decision entity, and the priorities of the policies, and the priorities of the M policies are different from each other.
  • the policy execution entity acquires the M policies, and the priority of the policy coordination dimension of each of the M policies, including:
  • the policy enforcement entity receives the M policies sent by the policy decision entity, and the priority of each execution action of the each policy, wherein the execution actions of the M policies have different priorities .
  • the policy coordination dimension of the mth policy is the mth policy Optimizing the target
  • the policy enforcement entity acquiring the M policies, and the priority of the policy coordination dimension of each of the M policies, including:
  • the policy execution entity receives the priority of the N optimization targets sent by the policy decision entity, where N is an integer greater than or equal to 1, and the priorities of the N optimization targets are different from each other;
  • the policy enforcement entity receives the M policies sent by the policy decision entity, and obtains an optimization target of each policy from each of the policies;
  • the policy execution entity determines a priority of the optimization target of the mth policy according to an optimization target of the mth policy and a priority of the N optimization targets.
  • the N optimization targets include any one or a combination of the following network service parameters: a network service NS instance The service load, the energy consumption of the NS instance, the virtual network function of the NS instance, the service route of the VNFFG, the service link congestion of the VNFFG of the NS instance, or the virtual network function VNF of the NS instance.
  • the method further includes:
  • the policy enforcement entity reports to the policy decision entity a policy that the execution conflict was not performed.
  • a second aspect of the present invention provides a policy coordination method in an NFV system, including:
  • the policy decision entity sets priorities for the policy coordination dimensions of the M policies, the M policies Each policy includes: a policy condition and an execution action, and the policy coordination dimension of the mth policy of the M policies includes at least one of the following information: the mth policy, the mth policy Performing an action or an optimization target of the m-th policy, wherein the optimization target of the m-th policy is an optimization target in a policy condition of the m-th policy, where M is an integer greater than or equal to 2, and a value of m Is 1, 2, ..., M;
  • the policy decision entity sends the priority of the M policies and the policy coordination dimensions of the M policies to the policy enforcement entity.
  • the method further includes:
  • the policy decision entity receives a policy reported by the policy execution entity that is not executed due to an execution conflict
  • the policy decision entity updates the priority of the policy coordination dimension of the policy that was not enforced due to the conflict.
  • a third aspect of the present invention provides a policy enforcement entity, including:
  • An obtaining module configured to acquire M policies, and priorities of each of the M policies in a policy coordination dimension, where each policy includes: a policy condition and an execution action, where the M policies are
  • the policy coordination dimension of the m policies includes at least one of the following information: the mth policy, an execution action of the mth policy, or an optimization target of the mth policy, optimization of the mth policy
  • the target is an optimization target in the policy condition of the mth policy, where M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M;
  • An execution module configured to: when the execution action of the first one of the M policies conflicts with the execution action of the second one of the M policies, according to the first policy, in the policy coordination dimension
  • the priority and the priority of the second policy under the policy coordination dimension determine a target execution action and perform the target execution action.
  • the executing module is specifically configured to:
  • the priority of the first policy in the policy coordination dimension is higher than the priority of the second policy in the policy coordination dimension, determining that the execution action of the first policy is the The target performs an action and performs an execution action of the first policy;
  • the priority of the first policy in the policy coordination dimension is lower than the priority of the second policy in the policy coordination dimension, determining that the execution action of the second policy is an action performed by the target And executing an execution action of the second policy.
  • the obtaining module is specifically configured to:
  • the acquisition module is specifically configured to:
  • the acquisition module is specifically used to:
  • the N optimization targets include any one or a combination of the following network service parameters: a network service NS instance The service load, the energy consumption of the NS instance, the virtual network function of the NS instance, the service route of the VNFFG, the service link congestion of the VNFFG of the NS instance, or the virtual network function VNF of the NS instance.
  • the method further includes:
  • the reporting module is configured to report to the policy decision entity a policy that is not executed because the execution conflict.
  • a fourth aspect of the present invention provides a policy decision entity, including:
  • a priority setting module configured to prioritize a policy coordination dimension of the M policies, each of the M policies includes: a policy condition and an execution action, and a policy coordination of the mth policy of the M policies
  • the dimension includes at least one of the following information: the mth policy, an execution action of the mth policy, or an optimization target of the mth policy, and an optimization target of the mth policy is the mth
  • the optimization goal in the policy condition of the strategy, M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M;
  • a sending module configured to send the priority of the M policies and the policy coordination dimensions of the M policies to the policy enforcement entity.
  • the policy decision entity further includes:
  • a receiving module configured to receive a policy that is not executed by the policy execution entity due to an execution conflict
  • a fifth aspect of the present invention provides a policy enforcement entity, including: a processor, a memory, a communication interface, and a system bus, wherein the memory and the communication interface are connected to the processor through the system bus and complete communication with each other
  • the memory is for storing computer execution instructions
  • the communication interface is for communicating with other devices
  • the processor is configured to execute the computer execution instructions to cause the policy execution entity to perform the first aspect and the A method according to any one of the first to sixth possible implementations of the invention.
  • a sixth aspect of the present invention provides a policy decision entity, including: a processor, a memory, a communication interface, and a system bus, wherein the memory and the communication interface are connected to the processor through the system bus and complete communication with each other
  • the memory is for storing computer execution instructions
  • the communication interface is for communicating with other devices
  • the processor is configured to execute the computer to execute instructions to cause the policy decision entity to perform the second aspect or the
  • a first possible implementation of the two aspects provides any of the methods described.
  • policy implementation The body acquires M policies, and the priority of the policy coordination dimension of each policy, and the policy coordination dimension of the mth policy of the M policies includes at least one of the following information: execution of the mth policy and the mth policy.
  • execution of the mth policy and the mth policy The action or the optimization goal of the mth policy, when the execution action of the first one of the M policies conflicts with the execution action of the second one of the M policies, the policy execution entity is in the policy coordination dimension according to the first policy
  • the lower priority and the priority of the second policy in the policy coordination dimension determine the target execution action and perform the target execution action.
  • the priority is set for the policy coordination dimension of each policy. When the two policies conflict, the priority of the policy is coordinated according to the policy of the policy, and the policy with the highest priority of the policy coordination dimension is executed to avoid the policy. Execution conflicts can lead to inconsistent executions that are expected by policy makers.
  • Figure 1 is a schematic diagram of the architecture of the management and orchestration functions of the NFV system
  • FIG. 2 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 1 of the present invention
  • FIG. 3 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 2 of the present invention
  • FIG. 4 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 3 of the present invention.
  • FIG. 5 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 4 of the present invention.
  • FIG. 6 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 5 of the present invention.
  • FIG. 7 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 6 of the present invention.
  • FIG. 8 is a schematic structural diagram of a policy execution entity according to Embodiment 7 of the present invention.
  • Embodiment 8 of the present invention is a schematic structural diagram of a policy decision entity according to Embodiment 8 of the present invention.
  • FIG. 10 is a schematic structural diagram of a policy execution entity according to Embodiment 9 of the present invention.
  • FIG. 11 is a schematic structural diagram of a policy decision entity according to Embodiment 10 of the present invention.
  • FIG. 1 is a schematic diagram of the structure of the management and orchestration function of the NFV system, and the NFV system includes an NFV orchestrator (NFV).
  • Orchestrator NFVO
  • VNFM Virtualized Network Function Manager
  • VIM Virtualized Infrastructure Manager
  • OSS Operation Support System
  • BSS business support System Support System
  • EMS Element Management System
  • VNF Virtualized Network Function
  • NFVI Network Function Virtualization Infrastructure
  • NFVO implements Network Service Descriptor (NSD), Virtualized Network Function Forwarding Graph (VNFFG) management, Network Service (NS) lifecycle management, and NFVO and VNFM work together to implement VNF lifecycle management and global view of resources.
  • NSD Network Service Descriptor
  • VNFFG Virtualized Network Function Forwarding Graph
  • NS Network Service lifecycle management
  • NFVO and VNFM work together to implement VNF lifecycle management and global view of resources.
  • the VNFM implements the lifecycle management of the VNF, including the management of the virtualized network function descriptor (VNFD), the creation of the VNF instance, the expansion/reduction of the VNF, the termination of the VNF instance, and the support for receiving the elasticity of the NFVO delivery. Scaling strategy to achieve flexible scaling of VNF.
  • VNFD virtualized network function descriptor
  • OSS/BSS is mainly composed of network management, system management, billing, business, accounting and customer service.
  • the systems are integrated through a unified information bus.
  • the EMS performs traditional fault management, configuration management, accounting management, performance management, and security management (Function Management, Accounting Management, Performance Management, Security Management, FCAPS) for the VNF.
  • the VNF corresponds to a physical network function (PNF) in a traditional non-virtualized network, such as a virtualized Evolved Packet Core (EPC) node, and the virtualized EPC node includes: a mobile management entity. (Mobile Management Entity, MME for short), Serving GateWay (SGW), and Packet Data Network GateWay (PGW). Among them, the functional behavior and state of the network are independent of whether the network is virtualized.
  • the VNF and PNF in the NFV system have the same functional behavior and external interfaces.
  • a VNF can be composed of multiple lower-level components. Therefore, a VNF can be deployed on multiple virtual machines (VMs), each of which carries a VNF component. VNF can also be deployed on a VM.
  • NFVI consists of hardware resources and virtual resources as well as virtual layers. From a VNF perspective, the virtualization layer and hardware resources appear to be an entity that can provide the required virtual resources.
  • NS Catalog Includes all NSDs that have been loaded.
  • VNF Catalog Includes all VNF Packages that have been loaded.
  • NFV Instances Includes all running NS instances and VNF instances.
  • Network Function Virtualization Infrastructure Resources NFVI Resources, including the status of all NFVI resources, such as available/reserved/allocated NFVI resources.
  • the OSS/BSS configures the automatic expansion/reduction strategy of the NS instance to the NFVO through the Os-Nfvo interface.
  • the automatic capacity expansion/reduction capacity policy indicates that the NFVO automatically performs the expansion/contraction operation of the NS under specific conditions. For example, when the performance indicator of the NS monitored by the NFVO "Calls-per-second" is greater than a preset threshold, the NFVO migrates the NS instance from the deployed template (Deployment Flavor) 1 to the deployed Template 2, wherein template 1 and template 2 differ in that one of the VNFs is expanded.
  • NFVO can also configure the automatic expansion/reduction strategy of the VNF instance to the VNFM through the Nfvo-Vnfm interface, instructing the VNFM to automatically perform the VNF expansion/reduction operation under certain conditions, and increase Or reduce the virtual resources of the VNF.
  • other lifecycle management policies are also configured in NFV Orchestrator or VNF Manager, such as self-healing strategies for self-healing functions of network business instances, load balancing policies for VNF instances, and so on. Multiple policies form a dive when executed in the same functional entity The execution conflict in.
  • two policies configured in NFVO determine the expansion operation of the VNF-A that constitutes the NS instance; and another policy requires the NFVO to monitor the NS instance.
  • the VNFFG when a virtual link (VL) in the VNFFG is blocked, the NFVO lowers the execution priority of the virtual link endpoint (VNF), and the virtual resource of the endpoint is moved out to other VNFs. Used, and VNF-A is an endpoint of the virtual link, and the shrinking operation is performed accordingly.
  • VNL virtual link
  • VNF-A is an endpoint of the virtual link, and the shrinking operation is performed accordingly.
  • FIG. 2 is a flowchart of a method for coordinating policies in an NFV system according to Embodiment 1 of the present invention, as shown in FIG.
  • the method provided in this embodiment may include the following steps:
  • Step 101 The policy execution entity acquires M policies, and the priority of each of the M policies in the policy coordination dimension, where each policy includes: a policy condition and an execution action, and a strategy of the mth policy among the M policies
  • the coordination dimension includes at least one of the following information: the mth strategy, the execution action of the mth policy, or the optimization goal of the mth policy, and the optimization goal of the mth strategy is the optimization goal in the policy condition of the mth policy
  • M is an integer greater than or equal to 2
  • the value of m is 1, 2, ..., M.
  • the policy decision entity may be an OSS/BSS or other device with a policy decision function.
  • the policy enforcement entity can be VNFVO or VNFM.
  • the policy enforcement entity obtains M policies, and the priorities of the M policies for the policy coordination dimension, specifically: the policy execution entity
  • M policies sent by the policy decision entity and the priorities of each policy are received, wherein the priorities of the M policies are different from each other.
  • the priority of each policy is set by the policy decision entity and sent to the policy enforcement entity.
  • each policy may include one or more execution actions.
  • different actions are performed.
  • an NS policy is represented as “a component object for the same monitoring. (eg VNF)
  • condition A is satisfied
  • action is performed
  • condition B is satisfied
  • action 2 is performed.
  • the policy decision entity is in addition to each policy sent to the policy execution entity.
  • the corresponding execution action is also added, that is, the priority is tied to the execution action of the policy. set.
  • the policy execution entity For each policy, if the policy coordination dimension of the mth policy is multiple execution actions of the mth policy, the policy execution entity acquires M policies, and each of the M policies has priority under the policy coordination dimension. Level, specifically: the policy execution entity receives the M policies sent by the policy decision entity, and the priority of each execution action of each policy, wherein the execution actions between the M policies have different priorities, but the same.
  • the multiple execution actions of the policy have the same priority, mainly because the individual policy conditions in the same policy cannot be satisfied at the same time. Therefore, the execution actions cannot be executed at the same time, and only one of the multiple execution actions is executed at a time. .
  • the policy enforcement entity acquires M policies, and the priority of each of the M policies in the policy coordination dimension, Specifically, the policy enforcement entity receives the priority of the N optimization targets sent by the policy decision entity, where N is an integer greater than or equal to 1, and the priorities of the N optimization targets are different from each other; then, the policy enforcement entity receives the policy decision The M policies sent by the entity obtain the optimization goal of each policy from each strategy. Finally, for each policy, the policy execution entity determines the first according to the optimization goal of the mth policy and the priority of the N optimization targets. The priority of the optimization goals of the m strategies.
  • the N optimization targets include any one or a combination of the following network service parameters: the traffic load of the NS instance, the energy consumption of the NS instance, the service routing of the VNFFG of the NS instance, or the traffic link congestion of the VNFFG of the NS instance.
  • the virtual network function VNF of the NS instance is faulty.
  • the policy execution entity may determine the priority of the optimization target according to the NS instance deployment requirement, the traffic volume size, or the QoS requirement, and deliver the priority of the optimization target to the policy execution entity.
  • the policy decision entity only needs to send the policy to the policy execution entity.
  • the policy execution entity obtains the optimization target of the policy from the received policy, and the policy execution entity delivers the optimization according to the policy decision entity.
  • the priority of the target is given the priority of the optimization goal of the strategy.
  • the priority of the optimization target may also be pre-configured to the policy execution entity instead of being delivered by the policy decision entity.
  • the priority of the policy under the policy coordination dimension is bound to the policy.
  • the policy coordination dimension is the optimization target, the policy is coordinated in the policy.
  • the priority under the dimension is not tied to the policy, only the best It is related to the importance of the implementation of the target.
  • Step 102 When the execution action of the first policy in the M policies conflicts with the execution action of the second policy in the M policies, the policy execution entity according to the priority and the second priority of the first policy in the policy coordination dimension The priority of the policy in the policy coordination dimension determines the target execution action and performs the target execution action.
  • the policy enforcement entity compares the priority of the first policy in the policy coordination dimension with the priority of the second policy in the policy coordination dimension; if the first policy has a higher priority in the policy coordination dimension than the second The priority of the policy in the policy coordination dimension, the policy execution entity determines that the execution action of the first policy is the target execution action, and performs the execution action of the first policy; if the first policy has a lower priority in the policy coordination dimension The priority of the second policy in the policy coordination dimension, the policy execution entity determines that the execution action of the second policy is the target execution action, and performs the execution action of the second policy.
  • the policy enforcement entity also reports to the policy decision entity a policy that the execution conflict is not executed. So that the policy decision entity adjusts the priority of each policy under the policy coordination dimension according to the priority of the policy coordination dimension of the policy that is not executed.
  • the policy enforcement entity acquires M policies and the priority of the policy coordination dimension of each policy, and the policy coordination dimension of the mth policy of the M policies includes at least one of the following information: the mth policy The execution action of the mth policy or the optimization goal of the mth policy, when the execution action of the first policy of the M policies conflicts with the execution action of the second policy of the M policies, the policy execution entity according to the The priority of a policy under the policy coordination dimension and the priority of the second policy under the policy coordination dimension determine a target execution action and perform a target execution action.
  • the priority is set for the policy coordination dimension of each policy. When the two policies conflict, the priority of the policy is coordinated according to the policy of the policy, and the policy with the highest priority of the policy coordination dimension is executed to avoid the policy. Execution conflicts can lead to inconsistent executions that are expected by policy makers.
  • FIG. 3 is a flowchart of a policy coordination method in an NFV system according to Embodiment 2 of the present invention. This embodiment is described from the perspective of a policy decision entity. As shown in FIG. 3, the method provided in this embodiment may include the following steps:
  • Step 201 The policy decision entity sets priorities for the policy coordination dimensions of the M policies.
  • Each of the M policies includes: a policy condition and an execution action, and the mth policy of the M policies
  • the slightly strategic coordination dimension includes at least one of the following information: the mth strategy, the execution action of the mth policy, or the optimization goal of the mth policy, and the optimization goal of the mth strategy is the policy condition of the mth policy
  • the optimization goal, M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M.
  • the policy coordination dimension may be set to the priority of the policy coordination dimension of the M policies according to the importance of the policy, and the priority of the policy coordination dimension of the high importance policy is set to be high. Priority, setting a low priority for the policy coordination dimension of a low-important policy to ensure that a highly important policy can be executed when two policies conflict.
  • the policy enforcement entity may set a priority for the optimization target according to the NS instance deployment requirement, the traffic volume size, or the QoS requirement.
  • Step 202 The policy decision entity sends the priority of the M policies and the policy coordination dimensions of the M policies to the policy execution entity.
  • the policy decision entity sets the priority in the policy coordination dimension for each policy, and sends each policy to the policy execution entity under the priority of the policy coordination dimension, and the policy execution entity is in the two
  • a policy conflicts, according to the priority of the policy in the policy coordination dimension, the policy with the highest priority of the policy coordination dimension is executed, and avoiding the execution conflict of the policy may result in an inconsistent execution effect with the policy maker.
  • FIG. 4 is a flowchart of a policy coordination method in the NFV system according to Embodiment 3 of the present invention.
  • the policy coordination dimension is used as the policy itself.
  • the method provided in this embodiment may include the following steps:
  • Step 301 The policy decision entity sends the priority of the auto-scale-policy 1 and the priority of the automatic expansion/reduction strategy 2 to the policy enforcement entity in the on-boarding phase.
  • the creation of NS consists of two phases: the NSD on-boarding phase and the create NS instance phase.
  • the policy decision entity may issue different policies and set different priorities for different policies.
  • the priority of the automatic capacity expansion/reduction policy delivered by the policy decision entity can be expressed as follows: NS on-boarding ((auto-scale-policy1, priority-x1), (auto-scale-policy2, priority-x2)) .
  • Step 302 The policy decision entity sends the priority of the runtime policy (runtime-policy-Info) 1 and the priority of the runtime policy 2 to the policy enforcement entity in the stage of creating the NS instance.
  • the priority of the runtime policy delivered by the policy decision entity can be expressed as follows: create NS instance ((runtime-policy-Info1, priority-y1), (runtime-policy-Info2, priority-y2)).
  • Step 303 The policy execution entity detects a policy execution conflict, determines a target policy according to a priority of the two policies in which the policy execution conflict occurs, and executes the target policy.
  • Step 304 The policy enforcement entity reports to the policy decision entity a policy that the policy execution conflict is not executed.
  • the policy determines the preemption execution according to the priority of the policy, and the preemptive failure policy is no longer executed, and the policy execution entity will An event notification is reported to the policy decision entity indicating that the policy was not executed due to a policy enforcement conflict.
  • FIG. 5 is a flowchart of a policy coordination method in an NFV system according to Embodiment 4 of the present invention.
  • an execution action of a policy coordination dimension is used as an example.
  • the embodiment provides The method can include the following steps:
  • Step 401 The policy decision entity sends the priority of the execution action 1 and the action 2 of the automatic expansion/reduction strategy 1 and the execution action 3 of the automatic capacity expansion/reduction strategy 2 to the policy execution entity in the NSD shelf-sending phase.
  • the automatic capacity expansion/reduction strategy 1 includes two execution actions: performing action 1 and performing action 2, and the automatic expansion/reduction strategy 2 includes an performing action: performing action 3.
  • the automatic capacity expansion/reduction strategy 1 if the number of calls per second of the NS instance exceeds the threshold Thresh1, the NS instance migrates from the template 1 to the template 2, that is, the VNF-A in the template 2 is scaled out; If the number of calls per second exceeds the threshold Thresh2 (Thresh2>Thresh1), the NS instance is migrated from template 1 to template 3, that is, the VNF-B in template 3 is expanded.
  • the automatic capacity expansion/reduction strategy 2 if a VL in the VNFFG of the NS instance is blocked, the VL endpoint VNF-B is subjected to a shrinking operation, and the resource is vacated to other VNFs.
  • the priority of the execution action of the automatic capacity expansion/reduction strategy delivered by the policy decision entity can be expressed as follows: NS on-boarding((auto-scale-policy1,(Action-x1,priority-x1)(Action-x2, Priority-x2), (auto-scale-policy2, (Action-x3, priority-x3)).
  • Step 402 The policy decision entity sends the priority of the execution action 1 of the runtime policy 1, the execution action 2 of the runtime policy 2, and the priority of the execution action 3 to the policy execution entity in the stage of creating the NS instance.
  • the priority of the execution action of the runtime policy delivered by the policy decision entity can be expressed as follows: NS on-boarding ((runtime-policy-Info1, (Action-y1, priority-y1), (runtime-policy-Info2, (Action-y2, priority-y2) (Action-y3, priority-y3)).
  • Step 403 The policy execution entity detects a policy execution conflict, determines a target policy according to a priority of an execution action of the two policies in which the policy execution conflict occurs, and executes the target policy.
  • Step 404 The policy enforcement entity reports to the policy decision entity a policy that the policy execution conflict is not performed.
  • the priority of the execution action of the policy determines which strategy preempts execution, and the preemptive failure policy does not. Then, the policy enforcement entity reports an event notification to the policy decision entity, indicating that the policy is not executed due to policy execution conflict.
  • FIG. 6 is a flowchart of a policy coordination method in an NFV system according to Embodiment 5 of the present invention.
  • an optimization target with a policy coordination dimension as a policy is taken as an example.
  • the method can include the following steps:
  • Step 501 The policy decision entity sends the priority of the optimization target 1 and the priority of the optimization target 2 to the policy execution entity in the NSD shelf stage.
  • the priority of the optimization target delivered by the policy decision entity can be expressed as follows: NS on-boarding ((policy-target-x1, priority-x1, (policy-target-x2, priority-x2).
  • Step 502 The policy decision entity sends the priority of the optimization target 3 and the priority of the optimization target 4 to the policy execution entity in the stage of creating the NS instance.
  • the priority of the optimization target delivered by the policy decision entity can be expressed as follows: create NS instance ((policy-target-y1, priority-y1, (policy-target-y2, priority-y2)).
  • Step 503 The policy decision entity sends a policy to the policy execution entity, where the policy includes an optimization target of the policy.
  • Step 504 The policy execution entity detects a policy execution conflict, determines a priority of the optimization target of the two policies in which the policy execution conflict occurs, determines a target policy according to a priority of the optimization target of the two policies in which the policy execution conflict occurs, and executes the target policy.
  • Step 505 The policy enforcement entity reports to the policy decision entity a policy that the policy execution conflict is not executed.
  • the priority of the optimization target of the policy determines which strategy preempts execution, and the preemptive failure strategy is not executed, and the policy execution entity An event notification is reported to the policy decision entity indicating that the policy was not executed due to a policy enforcement conflict.
  • FIG. 7 is a flowchart of a policy coordination method in an NFV system according to Embodiment 6 of the present invention. As shown in FIG. 7, the method provided in this embodiment may include the following steps:
  • Step 601 When the policy execution entity detects that two policies conflict, report the policy conflict policy to the policy decision entity.
  • Step 602 The policy decision entity determines the target policy according to the requirements of the operator's deployment, operation, or QoS guarantee.
  • Step 603 The policy decision entity sends the target policy to the policy execution entity.
  • Step 604 The policy execution entity executes the target policy.
  • a policy decision entity is used to coordinate the policy conflicts in a centralized manner, that is, when the policy execution entity detects a policy execution conflict, the conflicting policy is reported to the policy decision entity, and the policy decision entity is deployed, operated, or QoS according to the operator. Guarantee the requirements to determine the target strategy for preemptive execution, and return the result information of the coordinated policy to the policy enforcement entity.
  • One limitation of this embodiment is that if a policy decision entity interacts with multiple policy enforcement entities, the efficiency of the policy conflict coordination process depends on the processing power of the policy decision entity. When multiple policy enforcement entities report conflicting policies to the policy decision entity at the same time, the coordination judgment process within the policy decision entity often forms a processing bottleneck for the formation and distribution of policy coordination results.
  • FIG. 8 is a schematic structural diagram of a policy execution entity according to Embodiment 7 of the present invention. As shown in FIG. 8, the policy decision execution entity provided in this embodiment includes: an obtaining module 11 and an executing module 12.
  • the obtaining module 11 is configured to acquire M policies, and priorities of each of the M policies in a policy coordination dimension, where each policy includes: a policy condition and an execution action, and the M policies
  • the policy coordination dimension of the mth policy includes at least one of the following information: the mth policy, an execution action of the mth policy, or an optimization target of the mth policy, the mth
  • the optimization target of the strategy is an optimization target in the policy condition of the m-th policy, where M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M;
  • the executing module 12 is configured to: when the execution action of the first one of the M policies conflicts with the execution action of the second one of the M policies, the policy coordination dimension according to the first policy The lower priority and the priority of the second policy under the policy coordination dimension determine a target execution action and perform the target execution action.
  • the executing module 12 is specifically configured to: compare a priority of the first policy in the policy coordination dimension and a priority of the second policy in the policy coordination dimension; Determining the priority of the policy in the policy coordination dimension is higher than the priority of the second policy in the policy coordination dimension, determining that the execution action of the first policy is an action performed by the target, and performing the Performing an action of the first policy; if the priority of the first policy in the policy coordination dimension is lower than the priority of the second policy in the policy coordination dimension, determining execution of the second policy An action is performed as the target, and an execution action of the second policy is executed.
  • the acquiring module 11 is specifically configured to: receive the M policies sent by the policy decision entity, and the priority of each policy, The priorities of the M policies are different from each other.
  • the acquiring module 11 is specifically configured to: receive the M policies sent by the policy decision entity, and each of the foregoing The priority of each execution action of the policies, wherein the execution actions between the M policies have different priorities.
  • the acquiring module 11 is specifically configured to: receive the priority of the N optimization targets sent by the policy decision entity, where N is greater than or An integer equal to 1, the priorities of the N optimization targets are different from each other; and the M policies sent by the policy decision entity are received, and an optimization target of each policy is obtained from each policy; Determining an optimization target of the mth policy and a priority of the N optimization targets, and determining a priority of the optimization target of the mth policy.
  • the N optimization targets include any one of the following network service parameters or a combination thereof: a service load of the network service NS instance, an energy consumption of the NS instance, a virtual network function forwarding map of the NS instance, and a service routing of the VNFFG, NS The VNFFG service link congestion of the instance or the virtual network function VNF failure of the NS instance.
  • the policy decision entity further includes: a reporting module, configured to report, to the policy decision entity, a policy that is not executed because the execution conflict is performed.
  • a reporting module configured to report, to the policy decision entity, a policy that is not executed because the execution conflict is performed.
  • the policy enforcement entity provided in this embodiment may be used to perform the methods provided in the first embodiment and the third embodiment to the fifth embodiment.
  • the specific implementation manners and technical effects are similar, and details are not described herein again.
  • FIG. 9 is a schematic structural diagram of a policy decision entity according to Embodiment 8 of the present invention.
  • the policy decision entity provided in this embodiment includes: a priority determining module 21 and a sending module 22.
  • the priority setting module 21 is configured to set a priority for the policy coordination dimension of the M policies, where each of the M policies includes: a policy condition and an execution action, and a strategy of the mth policy of the M policies
  • the coordination dimension includes at least one of the following information: the mth policy, an execution action of the mth policy, or an optimization target of the mth policy, where the optimization target of the mth policy is the The optimization goal in the policy conditions of the m policies, M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M;
  • the sending module 22 is configured to send the priority of the M policies and the policy coordination dimensions of the M policies to the policy enforcement entity.
  • the policy decision entity may further include: a receiving module and an update module (not shown), and a receiving module, configured to receive a policy that is not executed by the policy execution entity due to an execution conflict; The priority of the policy coordination dimension of the policy that is not executed due to the conflict is updated.
  • the policy decision entity provided by this embodiment may be used to perform the method provided in the second embodiment.
  • the specific implementation manner and technical effects are similar, and details are not described herein again.
  • the policy decision entity 300 provided in this embodiment includes: a processor 31, a memory 32, a communication interface 33, and a system bus 34.
  • the memory 32 and the communication interface 33 are connected to the processor 31 via the system bus 34 for communication with each other, and the memory 32 is used to store computer execution instructions for use with other devices.
  • Communicating, the processor 31 is configured to run the computer to execute an instruction, causing the policy enforcement entity to perform the method as described below:
  • the policy includes: a policy condition and an execution action, where the M policies
  • the policy coordination dimension of the mth policy includes at least one of the following information: the mth policy, an execution action of the mth policy, or an optimization target of the mth policy, where the mth policy
  • the optimization target is an optimization target in the policy condition of the mth policy
  • M is an integer greater than or equal to 2
  • the value of m is 1, 2, ..., M;
  • the priority of the second policy in the policy coordination dimension determines a target execution action and performs the target execution action.
  • the processor 31 determines a target execution action according to a priority of the first policy in the policy coordination dimension and a priority of the second policy in the policy coordination dimension, and performs the target execution action, Specifically, comparing: a priority of the first policy in the policy coordination dimension and a priority of the second policy in the policy coordination dimension; if the first policy is prioritized in the policy coordination dimension If the level is higher than the priority of the second policy in the policy coordination dimension, determining that the execution action of the first policy is an action performed by the target, and performing an execution action of the first policy; Determining, by the first policy, that the priority of the second policy is lower than the priority of the second policy in the policy coordination dimension, determining that the execution action of the second policy is an action performed by the target, and executing The execution action of the second policy.
  • the processor 31 acquires the M policies, and the priority of the policy coordination dimension of each of the M policies, specifically: The M policies sent by the policy decision entity and the priorities of the policies are received by the communication interface 33, and the priorities of the M policies are different from each other.
  • the processor 31 acquires M policies, and a policy coordination dimension of each of the M policies
  • the priority is specifically: receiving, by the communication interface 33, the M policies sent by the policy decision entity, and the priority of each execution action of each of the policies, where the action between the M policies is performed
  • the priorities are different from each other.
  • the processor 31 acquires M policies, and priorities of policy coordination dimensions of each of the M policies Specifically, the priority of the N optimization targets sent by the policy decision entity is received through the communication interface 33, where N is an integer greater than or equal to 1, and the N optimization targets are optimized. The first levels are different from each other; and the M policies sent by the policy decision entity are received, and an optimization target of each of the policies is obtained from each of the policies; and then, according to the optimization goal of the mth policy And a priority of the N optimization targets, determining a priority of the optimization target of the mth policy.
  • the N optimization targets include any one or a combination of the following network service parameters: a service load of the network service NS instance, an energy consumption of the NS instance, and a virtual network function forwarding map of the NS instance.
  • the service link of the VNFFG of the NS instance is congested or the virtual network function of the NS instance is VNF.
  • the processor 31 is further configured to report, to the policy decision entity, a policy that is not performed because the execution conflict is performed.
  • the policy enforcement entity provided in this embodiment may be used to perform the methods provided in the first embodiment and the third embodiment to the fifth embodiment.
  • the specific implementation manners and technical effects are similar, and details are not described herein again.
  • FIG. 11 is a schematic structural diagram of a policy decision entity according to Embodiment 10 of the present invention.
  • the policy decision entity 400 provided in this embodiment includes: a processor 41, a memory 42, a communication interface 43, and a system bus 44.
  • the memory 42 and the communication interface 43 are connected to the processor 41 via the system bus 44 and perform communication with each other.
  • the memory 42 is used to store computer execution instructions, and the communication interface 43 is used for other devices.
  • Communicating, the processor 41 is configured to run the computer to execute an instruction, causing the policy decision entity to perform the method as described below:
  • each of the M policies Prioritizing the policy coordination dimensions of the M policies, each of the M policies includes: a policy condition and an execution action, where the policy coordination dimension of the mth policy includes at least one of the following information a: an m-th policy, an execution action of the m-th policy, or an optimization target of the m-th policy, where an optimization target of the m-th policy is in a policy condition of the m-th policy Optimization target, M is an integer greater than or equal to 2, and the value of m is 1, 2, ..., M;
  • the M policies and the priority of the policy coordination dimensions of the M policies are sent to the policy enforcement entity through the communication interface 43.
  • the processor 41 is further configured to receive, by using the communication interface 43, a policy that is not executed by the policy execution entity due to an execution conflict, and update a policy coordination dimension of the policy that is not executed because the conflict is not performed. priority.
  • the policy decision entity provided by this embodiment may be used to perform the method provided in the second embodiment.
  • the specific implementation manner and technical effects are similar, and details are not described herein again.
  • the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed.
  • the foregoing steps include the steps of the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Telephonic Communication Services (AREA)

Abstract

本发明实施例提供一种NFV系统中的策略协调方法和装置，策略执行实体获取M个策略以及策略执行冲突时的策略协调维度，以及每个策略在该策略协调维度下的优先级，当策略执行实体检测到两个策略发生冲突时，根据各策略在该策略协调维度下的优先级高低，执行优先级高的策略，避免策略的执行冲突会导致与策略制定者预期不一致的执行效果。

Description

NFV系统中的策略协调方法和装置 技术领域
本发明实施例涉及通信技术，尤其涉及一种网络功能虚拟化(Network Function Virtualization，简称NFV)系统中的策略协调方法和装置。
背景技术
NFV通过使用x86等通用性硬件以及虚拟化技术，来承载很多功能的软件处理，从而降低网络中昂贵的专用设备部署成本。NFV主要具有以下三个特性：1、将定义网络功能的软件从通用的硬件服务器、存储器以及网络交换机中完全分离出来，2、软件和硬件组件的独立的模块化特性，3、自动化的编排：基于通用硬件完全自动化地远程安装和管理软件功能。
NFV的管理和编排(Management and Orchestration，简称MANO)功能可以应用基于策略(Policy)的管理。例如：在网络业务(Network Service，简称NS)的生命周期管理中，运营支撑系统(Operations Support System，简称OSS)或业务支撑系统(Business Support System，简称BSS)通过Os-Nfvo接口向NFV编排器(NFV Orchestrator，简称NFVO)配置网络业务实例(NS Instance)的自动扩容/缩容策略(Auto-scale-policy)。该自动扩容/缩容策略指示NFVO在特定的条件下自动执行NS的扩容/缩容操作。相似地，在虚拟网络功能(Virtual Network Function，简称VNF)的生命周期管理中，NFVO同样可以通过Nfvo-Vnfm接口向VNF管理器(VNF Manager，简称VNFM)配置VNF实例的自动扩容/缩容策略，指示VNFM在特定的条件下自动执行VNF的扩容/缩容操作，增加或减少VNF的虚拟资源。
此外，在NFVO或VNFM中还会配置其他一些生命周期管理的策略，例如：对NS实例的自治愈功能配置的自治愈(self healing)策略，VNF实例的负载均衡策略等等。多个策略分别在不同的条件下产生，当多个策略在同一功能实体中执行时会形成潜在的执行冲突，不同策略的执行冲突会导致与策略制定者(例如：NFV运营商)预期不一致的执行效果。
发明内容
本发明实施例提供一种NFV系统中的策略协调方法和装置，能够避免策略的执行冲突会导致与策略制定者预期不一致的执行效果。
本发明第一方面提供一种NFV系统中的策略协调方法，包括：
策略执行实体获取M个策略，以及所述M个策略中每个策略在策略协调维度下的优先级，所述每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
当所述M个策略中的第一策略的执行动作与所述M个策略中的第二策略的执行动作发生冲突时，所述策略执行实体根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作。
结合第一方面，在第一方面的第一种可能的实现方式中，所述策略执行实体根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作，包括：
所述策略执行实体比较所述第一策略在所述策略协调维度的优先级和所述第二策略在所述策略协调维度下的优先级；
若所述第一策略在所述策略协调维度下的优先级高于所述第二策略在所述策略协调维度下的优先级，则所述策略执行实体确定所述第一策略的执行动作为所述目标执行动作，并执行所述第一策略的执行动作；
若所述第一策略在所述策略协调维度下的优先级低于所述第二策略在所述策略协调维度下的优先级，则所述策略执行实体确定所述第二策略的执行动作为所述目标执行动作，并执行所述第二策略的执行动作。
结合第一方面或第一方面的第一种可能的实现方式，在第一方面的第二种可能的实现方式中，若所述第m个的策略协调维度为所述第m个策略，则所述策略执行实体获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，包括：
所述策略执行实体接收策略决策实体发送的所述M个策略，以及所述每个策略的优先级，所述M个策略的优先级互不相同。
结合第一方面或第一方面的第一种可能的实现方式，在第一方面的第三种可能的实现方式中，若所述第m个策略的策略协调维度为所述第m个策略的多个执行动作，则所述策略执行实体获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，包括：
所述策略执行实体接收策略决策实体发送的所述M个策略，以及所述每个策略的每个执行动作的优先级，其中，所述M个策略之间的执行动作的优先级互不相同。
结合第一方面或第一方面的第一种可能的实现方式，在第一方面的第四种可能的实现方式中，若所述第m个策略的策略协调维度为所述第m个策略的优化目标，则所述策略执行实体获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，包括：
所述策略执行实体接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，所述N个优化目标的优先级互不相同；
所述策略执行实体接收所述策略决策实体发送的所述M个策略，从所述每个策略中获得所述每个策略的优化目标；
所述策略执行实体根据所述第m个策略的优化目标和所述N个优化目标的优先级，确定所述第m个策略的优化目标的优先级。
结合第一方面的第四种可能的实现方式，在第一方面的第五种可能的实现方式中，所述N个优化目标包括以下网络业务参数中的任意一个或其组合：网络业务NS实例的业务负载、NS实例的能耗、NS实例的虚拟网络功能转发图VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。
结合第一方面以及第一方面的第一种至第五种可能的实现方式中任意一种，在第一方面的第六种可能的实现方式中，所述方法还包括：
所述策略执行实体向策略决策实体上报因执行冲突没有被执行的策略。
本发明第二方面提供一种NFV系统中的策略协调方法，包括：
策略决策实体为M个策略的策略协调维度制定优先级，所述M个策 略中每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
所述策略决策实体将所述M个策略，以及所述M个策略的策略协调维度的优先级发送给策略执行实体。
结合第二方面，在第二方面的第一种可能的实现方式中，所述方法还包括：
所述策略决策实体接收所述策略执行实体上报的因执行冲突没有被执行的策略；
所述策略决策实体更新所述因冲突没有被执行的策略的策略协调维度的优先级。
本发明第三方面提供一种策略执行实体，包括：
获取模块，用于获取M个策略，以及所述M个策略中每个策略在策略协调维度下的优先级，所述每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
执行模块，用于当所述M个策略中的第一策略的执行动作与所述M个策略中的第二策略的执行动作发生冲突时，根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作。
结合第三方面，在第三方面的第一种可能的实现方式中，所述执行模块具体用于：
比较所述第一策略在所述策略协调维度的优先级和所述第二策略在所述策略协调维度下的优先级；
若所述第一策略在所述策略协调维度下的优先级高于所述第二策略在所述策略协调维度下的优先级，则确定所述第一策略的执行动作为所述 目标执行动作，并执行所述第一策略的执行动作；
若所述第一策略在所述策略协调维度下的优先级低于所述第二策略在所述策略协调维度下的优先级，则确定所述第二策略的执行动作为所述目标执行动作，并执行所述第二策略的执行动作。
结合第三方面或第三方面的第一种可能的实现方式，在第三方面的第二种可能的实现方式中，若所述第m个的策略协调维度为所述第m个策略，则所述获取模块具体用于：
接收策略决策实体发送的所述M个策略，以及所述每个策略的优先级，所述M个策略的优先级互不相同。
结合第三方面或第三方面的第一种可能的实现方式，在第三方面的第三种可能的实现方式中，若所述第m个策略的策略协调维度为所述第m个策略的多个执行动作，则所述获取模块具体用于：
接收策略决策实体发送的所述M个策略，以及所述每个策略的每个执行动作的优先级，其中，所述M个策略之间的执行动作的优先级互不相同。
结合第三方面或第三方面的第一种可能的实现方式，在第三方面的第四种可能的实现方式中，若所述第m个策略的策略协调维度为所述第m个策略的优化目标，则所述获取模块具体用于：
接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，所述N个优化目标的优先级互不相同；
接收所述策略决策实体发送的所述M个策略，从所述每个策略中获得所述每个策略的优化目标；
根据所述第m个策略的优化目标和所述N个优化目标的优先级，确定所述第m个策略的优化目标的优先级。
结合第三方面的第四种可能的实现方式，在第三方面的第五种可能的实现方式中，所述N个优化目标包括以下网络业务参数中的任意一个或其组合：网络业务NS实例的业务负载、NS实例的能耗、NS实例的虚拟网络功能转发图VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。
结合第三方面以及第三方面的第一种至第五种可能的实现方式中任 一一种，在第三方面的第六种可能的实现方式中，还包括：
上报模块，用于向策略决策实体上报因执行冲突没有被执行的策略。
本发明第四方面提供一种策略决策实体，包括：
优先级制定模块，用于为M个策略的策略协调维度制定优先级，所述M个策略中每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
发送模块，用于将所述M个策略，以及所述M个策略的策略协调维度的优先级发送给策略执行实体。
结合第四方面，在第四方面的第一种可能的实现方式中，所述策略决策实体还包括：
接收模块，用于接收所述策略执行实体上报的因执行冲突没有被执行的策略；
更新模块，用于更新所述因冲突没有被执行的策略的策略协调维度的优先级。
本发明第五方面提供一种策略执行实体，包括：处理器、存储器、通信接口和系统总线，所述存储器和所述通信接口通过所述系统总线与所述处理器连接并完成相互间的通信，所述存储器用于存储计算机执行指令，所述通信接口用于和其他设备进行通信，所述处理器用于运行所述计算机执行指令，使所述策略执行实体执行如本发明第一方面以及第一方面的第一种至第六种可能的实现方式提供的任一所述的方法。
本发明第六方面提供一种策略决策实体，包括：处理器、存储器、通信接口和系统总线，所述存储器和所述通信接口通过所述系统总线与所述处理器连接并完成相互间的通信，所述存储器用于存储计算机执行指令，所述通信接口用于和其他设备进行通信，所述处理器用于运行所述计算机执行指令，使所述策略决策实体执行如本发明第二方面或第二方面的第一种可能的实现方式提供的任一所述的方法。
本发明实施例提供的NFV系统中的策略协调方法和装置，策略执行实 体获取M个策略，以及每个策略的策略协调维度的优先级，M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：第m个策略、第m个策略的执行动作或者第m个策略的优化目标，当M个策略中的第一策略的执行动作与M个策略中的第二策略的执行动作发生冲突时，策略执行实体根据第一策略在该策略协调维度下的优先级和第二策略在该策略协调维度下的优先级确定目标执行动作，并执行目标执行动作。所述方法中，通过为每个策略的策略协调维度设置优先级，当两个策略发生冲突时，根据策略的策略协调维度优先级高低，执行策略协调维度的优先级高的策略，避免策略的执行冲突会导致与策略制定者预期不一致的执行效果。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为NFV系统的管理与编排功能的架构示意图；
图2为本发明实施例一提供的NFV系统中的策略协调方法的流程图；
图3为本发明实施例二提供的NFV系统中的策略协调方法的流程图；
图4为本发明实施例三提供的NFV系统中的策略协调方法的流程图；
图5为本发明实施例四提供的NFV系统中的策略协调方法的流程图；
图6为本发明实施例五提供的NFV系统中的策略协调方法的流程图；
图7为本发明实施例六提供的NFV系统中的策略协调方法的流程图；
图8为本发明实施例七提供的策略执行实体的结构示意图；
图9为本发明实施例八提供的策略决策实体的结构示意图；
图10为本发明实施例九提供的策略执行实体的结构示意图；
图11为本发明实施例十提供的策略决策实体的结构示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明实施例的方法应用在NFV系统，NFV系统具有自动编排与管理(Management and Orchestration，简称MAMO)功能，图1为NFV系统的管理与编排功能的架构示意图，NFV系统包括NFV编排器(NFV Orchestrator，简称NFVO)、虚拟网络功能管理器(Virtualized Network Function Manager，简称VNFM)、虚拟化基础设施管理器(Virtualized Infrastructure Manager，简称VIM)、运营支撑系统(Operation Support System，简称OSS)或业务支撑系统(Business Support System，简称BSS)、网关管理系统(Element Management System，简称EMS)、虚拟网络功能(Virtualized Network Function，简称VNF)、网络功能虚拟化基础设施(Network Function Virtualization Infrastructure，简称NFVI)。
其中，NFVO实现网络服务描述符(Network Service Descriptor，简称NSD)、虚拟网络功能转发图(Virtualized Network Function Forwarding Graph，简称VNFFG)的管理，网络服务(Network Service，简称NS)生命周期的管理，以及NFVO和VNFM配合实现VNF的生命周期管理和资源的全局视图功能。
VNFM实现VNF的生命周期管理，包括虚拟化网络功能描述符(Virtualized Infrastructure Manager，简称VNFD)的管理，VNF实例的创建，VNF的扩容/缩容，VNF实例的终止，支持接收NFVO下发的弹性伸缩(scaling)策略，实现VNF的弹性伸缩。
OSS/BSS主要由网络管理、系统管理、计费、营业、账务和客户服务等部分组成，系统间通过统一的信息总线有机整合在一起。
EMS针对VNF执行传统的故障管理、配置管理、计费管理、性能管理、安全管理(Fault Management，Configuration Management，Accounting Management，Performance Management，Security Management，简称FCAPS)功能。
VNF对应于传统非虚拟化网络中的物理网络功能(Physical Network Function，简称PNF)，如虚拟化的演进分组核心网(Evolved Packet Core，简称EPC)节点，虚拟化的EPC节点包括：移动管理实体(Mobile Management Entity，简称MME)、服务网关(Serving GateWay，简称SGW)和分组数据网关(Packet Data Network GateWay，简称PGW)等。其中，网络的功能性行为和状态与网络是否虚拟化无关，NFV系统中VNF和PNF拥有相同的功能性行为和外部接口。VNF可以由多个更低级别的组件来组成，因此，一个VNF可以部署在多个虚拟机(Virtualized Machine，简称VM)上，每个VM承载一个VNF组件。VNF也可以部署在一个VM上。NFVI由硬件资源和虚拟资源以及虚拟层组成，从VNF的角度来说，虚拟化层和硬件资源看起来是一个能够提供所需虚拟资源的实体。
MANO的数据都存放在下面四个储存库中：网络业务目录(NS Catalog)：包括所有已经加载的NSD。虚拟网络功能目录(VNF Catalog)：包括所有已经加载的VNF Packages。网络功能虚拟化实例(NFV Instances)：包括所有正在运行的NS实例和VNF实例。网络功能虚拟化基础设施资源(NFVI Resources)，包括所有NFVI资源的状态，例如，可用的/已预留的/已分配的NFVI资源。
MANO功能可以应用基于策略的管理。例如：在NS的生命周期管理中，OSS/BSS通过Os-Nfvo接口向NFVO配置NS实例的自动扩容/缩容策略。该自动扩容/缩容策略指示NFVO在特定的条件下自动执行NS的扩容/缩容操作。例如，当NFVO所监视的NS的性能指标“每秒呼叫数(Calls-per-second)”大于一个预先设置的门限时，NFVO将该NS实例从部署的模板(Deployment Flavor)1迁移到部署的模板2，其中模板1和模板2的差别在于对其中的一个VNF进行了扩容操作。
相似地，在VNF的生命周期管理中，NFVO同样可以通过Nfvo-Vnfm接口向VNFM配置VNF实例的自动扩容/缩容策略，指示VNFM在特定的条件下自动执行VNF的扩容/缩容操作，增加或减少VNF的虚拟资源。此外，在NFV Orchestrator或VNF Manager中还会配置其他一些生命周期管理的策略，例如：对网络业务实例的自治愈功能配置的自治愈策略，VNF实例的负载均衡策略等等。多个策略在相同的功能实体中执行时会形成潜 在的执行冲突。例如，在NFVO中配置的两个策略，其中之一是监视NS实例的每秒呼叫数达到一个门限时，确定对组成NS实例的VNF-A实施扩容操作；而另一个策略要求NFVO监视NS实例的VNFFG，当VNFFG中的某条虚拟链路(Virtual Link，简称VL)阻塞时，NFVO对该虚拟链路的端点(VNF)降低执行优先级，将该端点的虚拟资源腾挪出来给其他的VNF使用，而VNF-A是该虚拟链路的一个端点，相应地执行缩容操作。两个策略在不同的条件下产生了对同一个VNF-A执行扩容/缩容两个效果相反的操作，从而形成了策略执行冲突。
为了解决现有技术的问题，本发明实施例一提供一种NFV系统中的策略协调方法，图2为本发明实施例一提供的NFV系统中的策略协调方法的流程图，如图2所示，本实施例提供的方法可以包括以下步骤：
步骤101、策略执行实体获取M个策略，以及M个策略中每个策略在策略协调维度下的优先级，每个策略包括：策略条件和执行动作，M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：第m个策略、第m个策略的执行动作或者第m个策略的优化目标，第m个策略的优化目标为第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M。
本实施例中，策略决策实体可以为OSS/BSS或者其他具有策略决策功能的设备。策略执行实体可以为VNFVO或VNFM。
针对每个策略，若第m个策略的策略协调维度为第m个策略本身，则策略执行实体获取M个策略，以及M个策略中针对该策略协调维度的优先级，具体为：策略执行实体接收策略决策实体发送的M个策略，以及每个策略的优先级，其中，M个策略的优先级互不不同。每个策略的优先级由策略决策实体设置并下发给策略执行实体。
本实施例中，每个策略可以包括一个或多个执行动作(action)，当该策略满足不同的策略条件时，执行不同的动作，例如，一个NS策略表示为“对于同一个监控的组成对象(例如VNF)，如果满足条件A，则执行动作，如果满足条件B，则执行动作2。对于这种具有多个执行动作的策略，策略决策实体在下发给策略执行实体的每个策略中除了配以确定的执行优先级外，还增加相应的执行动作，即将优先级与策略的执行动作绑 定。
针对每个策略，若第m个策略的策略协调维度为第m个策略的多个执行动作，则策略执行实体获取M个策略，以及M个策略中每个策略在该策略协调维度下的优先级，具体为：策略执行实体接收策略决策实体发送的M个策略，以及每个策略的每个执行动作的优先级，其中，M个策略之间的执行动作的优先级不同，但是，同一个策略的多个执行动作的优先级相同，主要是因为同一个策略中的各个策略条件不可能同时满足，因此，执行动作不可能同时执行，每次只会执行多个执行动作中的一个执行动作。
针对每个策略，若第m个策略的策略协调维度为第m个策略的优化目标，则策略执行实体获取M个策略，以及M个策略中每个策略在该策略协调维度下的优先级，具体为：首先，策略执行实体接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，N个优化目标的优先级互不相同；然后，策略执行实体接收策略决策实体发送的M个策略，从每个策略中获得每个策略的优化目标；最后，针对每个策略，策略执行实体根据第m个策略的优化目标和该N个优化目标的优先级，确定第m个策略的优化目标的优先级。
其中，N个优化目标包括以下网络业务参数中的任意一个或其组合：NS实例的业务负载、NS实例的能耗、NS实例的VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。策略执行实体可以根据NS实例部署需求、业务量规模或QoS需求确定优化目标的优先级，并将优化目标的优先级的下发给策略执行实体。后续，在策略创建过程中，策略决策实体仅需要将策略下发给策略执行实体，策略执行实体从接收到的该策略中获得该策略的优化目标，策略执行实体根据策略决策实体下发的优化目标的优先级，得到给策略的优化目标的优先级。本实施例中，优化目标的优先级也可以预先配置到策略执行实体，而不是由策略决策实体下发。
通过上述描述可知，当策略协调维度为策略本身或策略中的执行动作时，策略在该策略协调维度下的优先级与策略绑定，当策略协调维度为优化目标时，该策略在该策略协调维度下的优先级并不与策略绑定，只与优 化目标执行的重要程度有关。
步骤102、当M个策略中的第一策略的执行动作与M个策略中的第二策略的执行动作发生冲突时，策略执行实体根据第一策略在该策略协调维度下的优先级和第二策略在该策略协调维度下的优先级确定目标执行动作，并执行目标执行动作。
具体的，策略执行实体比较第一策略在该策略协调维度下的优先级和第二策略在该策略协调维度下的优先级；若第一策略在该策略协调维度下的优先级高于第二策略在该策略协调维度下的优先级，则策略执行实体确定第一策略的执行动作为目标执行动作，并执行第一策略的执行动作；若第一策略在该策略协调维度的优先级低于第二策略在该策略协调维度下的优先级，则策略执行实体确定第二策略的执行动作为目标执行动作，并执行第二策略的执行动作。
可选的，策略执行实体还向策略决策实体上报因执行冲突没有被执行的策略。以便策略决策实体根据该没有被执行的策略的策略协调维度的优先级调整各个策略在该策略协调维度下的优先级。
本实施例中，策略执行实体获取M个策略，以及每个策略的策略协调维度的优先级，M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：第m个策略、第m个策略的执行动作或者第m个策略的优化目标，当M个策略中的第一策略的执行动作与M个策略中的第二策略的执行动作发生冲突时，策略执行实体根据第一策略在该策略协调维度下的优先级和第二策略在该策略协调维度下的优先级确定目标执行动作，并执行目标执行动作。所述方法中，通过为每个策略的策略协调维度设置优先级，当两个策略发生冲突时，根据策略的策略协调维度优先级高低，执行策略协调维度的优先级高的策略，避免策略的执行冲突会导致与策略制定者预期不一致的执行效果。
图3为本发明实施例二提供的NFV系统中的策略协调方法的流程图，本实施例从策略决策实体的角度描述，如图3所示，本实施例提供的方法可以包括以下步骤：
步骤201、策略决策实体为M个策略的策略协调维度制定优先级，M个策略中每个策略包括：策略条件和执行动作，M个策略中的第m个策 略的策略协调维度包括以下信息中的至少一个：第m个策略、第m个策略的执行动作或者第m个策略的优化目标，第m个策略的优化目标为第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M。
策略决策实体在为M个策略的策略协调维度制定优先级时，可以根据策略的重要性为M个策略的策略协调维度设置优先级，为重要性高的策略的策略协调维度设置优先级设置高优先级，为重要性低的策略的策略协调维度制定优先级设置低优先级，以保证当两个策略发生冲突时，重要性高的策略能够被执行。当策略协调维度为优化目标时，策略执行实体可以根据NS实例部署需求、业务量规模或QoS需求为优化目标设置优先级。
步骤202、策略决策实体将M个策略，以及M个策略的策略协调维度的优先级发送给策略执行实体。
本实施例中，策略决策实体通过为每个策略设置在策略协调维度下的优先级，并将每个策略在该策略协调维度下的优先级下发给策略执行实体，以策略执行实体在两个策略发生冲突时，根据策略在该策略协调维度下的优先级高低，执行策略协调维度的优先级高的策略，避免策略的执行冲突会导致与策略制定者预期不一致的执行效果。
以下将通过具体的实施例对实施例一的方案进行详细说明，图4为本发明实施例三提供的NFV系统中的策略协调方法的流程图，本实施例中以策略协调维度为策略本身为例进行说明，如图4所示，本实施例提供的方法可以包括以下步骤：
步骤301、策略决策实体在NSD上架(on-boarding)阶段向策略执行实体发送自动扩容/缩容策略1(auto-scale-policy)的优先级、自动扩容/缩容策略2的优先级。
NS的创建包括两个阶段：NSD on-boarding阶段和创建NS实例(create NS instance)阶段，在两个阶段策略决策实体可能下发不同的策略，并为不同的策略设置不同的优先级。策略决策实体下发的自动扩容/缩容策略的优先级可以通过如下方式表示：NS on-boarding((auto-scale-policy1,priority-x1),(auto-scale-policy2,priority-x2))。
步骤302、策略决策实体在创建NS实例阶段向策略执行实体发送运行时策略(runtime-policy-Info)1的优先级、运行时策略2的优先级。
策略决策实体下发的运行时策略的优先级可以通过如下方式表示：create NS instance((runtime-policy-Info1,priority-y1),(runtime-policy-Info2,priority-y2))。
步骤303、策略执行实体检测策略执行冲突，根据发生策略执行冲突的两个策略的优先级确定目标策略，执行目标策略。
步骤304、策略执行实体向策略决策实体上报因策略执行冲突没有被执行的策略。
本实施例中，通过为每个策略设置唯一的优先级，当两个策略的执行动作相冲突时，根据策略的优先级决定哪个策略抢占执行，抢占失败的策略不再执行，策略执行实体会向策略决策实体上报一个事件通知，指示该策略由于策略执行冲突而未被执行。
图5为本发明实施例四提供的NFV系统中的策略协调方法的流程图，本实施例中以策略协调维度为策略的执行动作为例进行说明，如图5所示，本实施例提供的方法可以包括以下步骤：
步骤401、策略决策实体在NSD上架阶段向策略执行实体发送自动扩容/缩容策略1的执行动作1和执行动作2的优先级、自动扩容/缩容策略2的执行动作3的优先级。
本实施例中，自动扩容/缩容策略1包括两个执行动作：执行动作1和执行动作2，自动扩容/缩容策略2包括一个执行动作：执行动作3。自动扩容/缩容策略1中，若NS实例的每秒呼叫数超过门限Thresh1时，NS实例从模板1迁移到模板2上，即对模板2中的VNF-A进行扩容(scale out)操作；若每秒呼叫数超过门限Thresh2(Thresh2>Thresh1)时，NS实例从模板1迁移到模板3上，即对模板3中的VNF-B进行扩容操作。自动扩容/缩容策略2中，如果NS实例的VNFFG中某一VL阻塞时，则对VL端点VNF-B进行缩容操作，腾挪出资源给其他VNF。
策略决策实体下发的自动扩容/缩容策略的执行动作的优先级可以通过如下方式表示：NS on-boarding((auto-scale-policy1,(Action-x1,priority-x1)(Action-x2,priority-x2),(auto-scale-policy2，(Action-x3，priority-x3))。
步骤402、策略决策实体在创建NS实例阶段向策略执行实体发送运行时策略1的执行动作1的优先级、运行时策略2的执行动作2和执行动作3的优先级。
策略决策实体下发的运行时策略的执行动作的优先级可以通过如下方式表示：NS on-boarding((runtime-policy-Info1,(Action-y1,priority-y1),(runtime-policy-Info2，(Action-y2,priority-y2)(Action-y3，priority-y3))。
步骤403、策略执行实体检测策略执行冲突，根据发生策略执行冲突的两个策略的执行动作的优先级确定目标策略，执行目标策略。
步骤404、策略执行实体向策略决策实体上报因策略执行冲突没有被执行的策略。
本实施例中，通过为每个策略的各执行动作设置唯一的优先级，当两个策略的执行动作相冲突时，根据策略的执行动作的优先级决定哪个策略抢占执行，抢占失败的策略不再执行，策略执行实体会向策略决策实体上报一个事件通知，指示该策略由于策略执行冲突而未被执行。
图6为本发明实施例五提供的NFV系统中的策略协调方法的流程图，本实施例中以策略协调维度为策略的优化目标为例进行说明，如图6所示，本实施例提供的方法可以包括以下步骤：
步骤501、策略决策实体在NSD上架阶段向策略执行实体发送优化目标1的优先级、优化目标2的优先级。
策略决策实体下发的优化目标的优先级可以通过如下方式表示：NS on-boarding((policy-target-x1,priority-x1,(policy-target-x2,priority-x2)。
步骤502、策略决策实体在创建NS实例阶段向策略执行实体发送优化目标3的优先级、优化目标4的优先级。
策略决策实体下发的优化目标的优先级可通过如下方式表示：create NS instance((policy-target-y1,priority-y1,(policy-target-y2,priority-y2)。
步骤503、策略决策实体向策略执行实体发送策略，该策略中包括策略的优化目标。
步骤504、策略执行实体检测策略执行冲突，确定发生策略执行冲突的两个策略的优化目标的优先级，根据发生策略执行冲突的两个策略的优化目标的优先级确定目标策略，执行目标策略。
步骤505、策略执行实体向策略决策实体上报因策略执行冲突没有被执行的策略。
本实施例中，通过为优化目标设置优先级，当两个策略的执行动作相冲突时，根据策略的优化目标的优先级决定哪个策略抢占执行，抢占失败的策略不再执行，策略执行实体会向策略决策实体上报一个事件通知，指示该策略由于策略执行冲突而未被执行。
图7为本发明实施例六提供的NFV系统中的策略协调方法的流程图，如图7所示，本实施例提供的方法可以包括以下步骤：
步骤601、当策略执行实体检测到两个策略发生冲突时，向策略决策实体上报发生策略冲突的策略。
步骤602、策略决策实体根据运营商的部署、运营或QoS保障等需求确定目标策略。
步骤603、策略决策实体向策略执行实体发送目标策略。
步骤604、策略执行实体执行目标策略。
本实施例采取策略决策实体集中协调策略冲突的方法，即：当策略执行实体检测到策略执行冲突时，将冲突的策略上报给策略决策实体，由策略决策实体根据运营商的部署、运营或QoS保障等需求来确定抢占执行的目标策略，并将策略协调后的结果信息返回给策略执行实体。本实施例的一个局限性表现在：如果一个策略决策实体和多个策略执行实体交互时，策略冲突协调过程的效率取决于策略决策实体的处理能力。当多个策略执行实体同时向策略决策实体上报需要协调的冲突策略时，策略决策实体内部的协调判断过程往往会形成策略协调结果形成和分发的处理瓶颈。
图8为本发明实施例七提供的策略执行实体的结构示意图，如图8所示，本实施例提供的策略决策执行实体包括：获取模块11和执行模块12。
其中，获取模块11，用于获取M个策略，以及所述M个策略中每个策略在策略协调维度下的优先级，所述每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……， M；
执行模块12，用于当所述M个策略中的第一策略的执行动作与所述M个策略中的第二策略的执行动作发生冲突时，根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作。
可选的，所述执行模块12具体用于：比较所述第一策略在所述策略协调维度的优先级和所述第二策略在所述策略协调维度下的优先级；若所述第一策略在所述策略协调维度下的优先级高于所述第二策略在所述策略协调维度下的优先级，则确定所述第一策略的执行动作为所述目标执行动作，并执行所述第一策略的执行动作；若所述第一策略在所述策略协调维度下的优先级低于所述第二策略在所述策略协调维度下的优先级，则确定所述第二策略的执行动作为所述目标执行动作，并执行所述第二策略的执行动作。
若所述第m个的策略协调维度为所述第m个策略，则所述获取模块11具体用于：接收策略决策实体发送的所述M个策略，以及所述每个策略的优先级，所述M个策略的优先级互不相同。
若所述第m个策略的策略协调维度为所述第m个策略的多个执行动作，则所述获取模块11具体用于：接收策略决策实体发送的所述M个策略，以及所述每个策略的每个执行动作的优先级，其中，所述M个策略之间的执行动作的优先级互不相同。
若所述第m个策略的策略协调维度为所述第m个策略的优化目标，则所述获取模块11具体用于：接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，所述N个优化目标的优先级互不相同；接收所述策略决策实体发送的所述M个策略，从所述每个策略中获得所述每个策略的优化目标；根据所述第m个策略的优化目标和所述N个优化目标的优先级，确定所述第m个策略的优化目标的优先级。
其中，所述N个优化目标包括以下网络业务参数中的任意一个或其组合：网络业务NS实例的业务负载、NS实例的能耗、NS实例的虚拟网络功能转发图VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。
可选的，策略决策实体还包括：上报模块，用于向策略决策实体上报因执行冲突没有被执行的策略。
本实施例提供的策略执行实体，可用于执行实施例一、实施例三至实施例五提供的方法，具体实现方式和技术效果类似，这里不再赘述。
图9为本发明实施例八提供的策略决策实体的结构示意图，如图9所示，本实施例提供的策略决策实体包括：优先级制定模块21和发送模块22。
优先级制定模块21，用于为M个策略的策略协调维度制定优先级，所述M个策略中每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
发送模块22，用于将所述M个策略，以及所述M个策略的策略协调维度的优先级发送给策略执行实体。
可选的，策略决策实体还可以包括：接收模块和更新模块(图中未示出)，接收模块，用于接收所述策略执行实体上报的因执行冲突没有被执行的策略；更新模块，用于更新所述因冲突没有被执行的策略的策略协调维度的优先级。
本实施例提供的策略决策实体，可用于执行实施例二提供的方法，具体实现方式和技术效果类似，这里不再赘述。
图10为本发明实施例九提供的策略执行实体的结构示意图，如图10所示，本实施例提供的策略决策实体300包括：处理器31、存储器32、通信接口33和系统总线34，所述存储器32和所述通信接口33通过所述系统总线34与所述处理器31连接并完成相互间的通信，所述存储器32用于存储计算机执行指令，所述通信接口33用于和其他设备进行通信，所述处理器31用于运行所述计算机执行指令，使所述策略执行实体执行如下所述的方法：
获取M个策略，以及所述M个策略中每个策略在策略协调维度下的优先级，所述每个策略包括：策略条件和执行动作，所述M个策略中的 第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
当所述M个策略中的第一策略的执行动作与所述M个策略中的第二策略的执行动作发生冲突时，根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作。
所述处理器31根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作，具体为：比较所述第一策略在所述策略协调维度的优先级和所述第二策略在所述策略协调维度下的优先级；若所述第一策略在所述策略协调维度下的优先级高于所述第二策略在所述策略协调维度下的优先级，则确定所述第一策略的执行动作为所述目标执行动作，并执行所述第一策略的执行动作；若所述第一策略在所述策略协调维度下的优先级低于所述第二策略在所述策略协调维度下的优先级，则确定所述第二策略的执行动作为所述目标执行动作，并执行所述第二策略的执行动作。
若所述第m个的策略协调维度为所述第m个策略，则所述处理器31获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，具体为：通过通信接口33接收策略决策实体发送的所述M个策略，以及所述每个策略的优先级，所述M个策略的优先级互不相同。
若所述第m个策略的策略协调维度为所述第m个策略的多个执行动作，则所述处理器31获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，具体为：通过通信接口33接收策略决策实体发送的所述M个策略，以及所述每个策略的每个执行动作的优先级，其中，所述M个策略之间的执行动作的优先级互不相同。
若所述第m个策略的策略协调维度为所述第m个策略的优化目标，则所述处理器31获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，具体为：通过通信接口33接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，所述N个优化目标的优 先级互不相同；以及接收所述策略决策实体发送的所述M个策略，从所述每个策略中获得所述每个策略的优化目标；然后，根据所述第m个策略的优化目标和所述N个优化目标的优先级，确定所述第m个策略的优化目标的优先级。
本实施例中，所述N个优化目标包括以下网络业务参数中的任意一个或其组合：网络业务NS实例的业务负载、NS实例的能耗、NS实例的虚拟网络功能转发图VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。
可选的，所述处理器31还用于：向策略决策实体上报因执行冲突没有被执行的策略。
本实施例提供的策略执行实体，可用于执行实施例一、实施例三至实施例五提供的方法，具体实现方式和技术效果类似，这里不再赘述。
图11为本发明实施例十提供的策略决策实体的结构示意图，如图11所示，本实施例提供的策略决策实体400包括：处理器41、存储器42、通信接口43和系统总线44，所述存储器42和所述通信接口43通过所述系统总线44与所述处理器41连接并完成相互间的通信，所述存储器42用于存储计算机执行指令，所述通信接口43用于和其他设备进行通信，所述处理器41用于运行所述计算机执行指令，使所述策略决策实体执行如下所述的方法：
为M个策略的策略协调维度制定优先级，所述M个策略中每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
通过所述通信接口43将所述M个策略，以及所述M个策略的策略协调维度的优先级发送给策略执行实体。
可选的，所述处理器41还用于通过所述通信接口43接收所述策略执行实体上报的因执行冲突没有被执行的策略，更新所述因冲突没有被执行的策略的策略协调维度的优先级。
本实施例提供的策略决策实体，可用于执行实施例二提供的方法，具体实现方式和技术效果类似，这里不再赘述。
本领域普通技术人员可以理解：实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成，前述的程序可以存储于一计算机可读取存储介质中，该程序在执行时，执行包括上述方法实施例的步骤；而前述的存储介质包括：ROM、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (20)

  1. 一种NFV系统中的策略协调方法，其特征在于，包括：
    策略执行实体获取M个策略，以及所述M个策略中每个策略在策略协调维度下的优先级，所述每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
    当所述M个策略中的第一策略的执行动作与所述M个策略中的第二策略的执行动作发生冲突时，所述策略执行实体根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作。
  2. 根据权利要求1所述的方法，其特征在于，所述策略执行实体根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作，包括：
    所述策略执行实体比较所述第一策略在所述策略协调维度的优先级和所述第二策略在所述策略协调维度下的优先级；
    若所述第一策略在所述策略协调维度下的优先级高于所述第二策略在所述策略协调维度下的优先级，则所述策略执行实体确定所述第一策略的执行动作为所述目标执行动作，并执行所述第一策略的执行动作；
    若所述第一策略在所述策略协调维度下的优先级低于所述第二策略在所述策略协调维度下的优先级，则所述策略执行实体确定所述第二策略的执行动作为所述目标执行动作，并执行所述第二策略的执行动作。
  3. 根据权利要求1或2所述的方法，其特征在于，若所述第m个的策略协调维度为所述第m个策略，则所述策略执行实体获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，包括：
    所述策略执行实体接收策略决策实体发送的所述M个策略，以及所述每个策略的优先级，所述M个策略的优先级互不相同。
  4. 根据权利要求1或2所述的方法，其特征在于，若所述第m个策 略的策略协调维度为所述第m个策略的多个执行动作，则所述策略执行实体获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，包括：
    所述策略执行实体接收策略决策实体发送的所述M个策略，以及所述每个策略的每个执行动作的优先级，其中，所述M个策略之间的执行动作的优先级互不相同。
  5. 根据权利要求1或2所述的方法，其特征在于，若所述第m个策略的策略协调维度为所述第m个策略的优化目标，则所述策略执行实体获取M个策略，以及所述M个策略中每个策略的策略协调维度的优先级，包括：
    所述策略执行实体接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，所述N个优化目标的优先级互不相同；
    所述策略执行实体接收所述策略决策实体发送的所述M个策略，从所述每个策略中获得所述每个策略的优化目标；
    所述策略执行实体根据所述第m个策略的优化目标和所述N个优化目标的优先级，确定所述第m个策略的优化目标的优先级。
  6. 根据权利要求5所述的方法，其特征在于，所述N个优化目标包括以下网络业务参数中的任意一个或其组合：网络业务NS实例的业务负载、NS实例的能耗、NS实例的虚拟网络功能转发图VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。
  7. 根据权利要求1-6中任一项所述的方法，其特征在于，所述方法还包括：
    所述策略执行实体向策略决策实体上报因执行冲突没有被执行的策略。
  8. 一种NFV系统中的策略协调方法，其特征在于，包括：
    策略决策实体为M个策略的策略协调维度制定优先级，所述M个策略中每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略 的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
    所述策略决策实体将所述M个策略，以及所述M个策略的策略协调维度的优先级发送给策略执行实体。
  9. 根据权利要求8所述的方法，其特征在于，所述方法还包括：
    所述策略决策实体接收所述策略执行实体上报的因执行冲突没有被执行的策略；
    所述策略决策实体更新所述因冲突没有被执行的策略的策略协调维度的优先级。
  10. 一种策略执行实体，其特征在于，包括：
    获取模块，用于获取M个策略，以及所述M个策略中每个策略在策略协调维度下的优先级，所述每个策略包括：策略条件和执行动作，所述M个策略中的第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
    执行模块，用于当所述M个策略中的第一策略的执行动作与所述M个策略中的第二策略的执行动作发生冲突时，根据所述第一策略在所述策略协调维度下的优先级和所述第二策略在所述策略协调维度下的优先级确定目标执行动作，并执行所述目标执行动作。
  11. 根据权利要求10所述的策略执行实体，其特征在于，所述执行模块具体用于：
    比较所述第一策略在所述策略协调维度的优先级和所述第二策略在所述策略协调维度下的优先级；
    若所述第一策略在所述策略协调维度下的优先级高于所述第二策略在所述策略协调维度下的优先级，则确定所述第一策略的执行动作为所述目标执行动作，并执行所述第一策略的执行动作；
    若所述第一策略在所述策略协调维度下的优先级低于所述第二策略在所述策略协调维度下的优先级，则确定所述第二策略的执行动作为所述目标执行动作，并执行所述第二策略的执行动作。
  12. 根据权利要求10或11所述的策略执行实体，其特征在于，若所述第m个的策略协调维度为所述第m个策略，则所述获取模块具体用于：
    接收策略决策实体发送的所述M个策略，以及所述每个策略的优先级，所述M个策略的优先级互不相同。
  13. 根据权利要求10或11所述的策略执行实体，其特征在于，若所述第m个策略的策略协调维度为所述第m个策略的多个执行动作，则所述获取模块具体用于：
    接收策略决策实体发送的所述M个策略，以及所述每个策略的每个执行动作的优先级，其中，所述M个策略之间的执行动作的优先级互不相同。
  14. 根据权利要求10或11所述的策略执行实体，其特征在于，若所述第m个策略的策略协调维度为所述第m个策略的优化目标，则所述获取模块具体用于：
    接收策略决策实体发送的N个优化目标的优先级，N为大于或等于1的整数，所述N个优化目标的优先级互不相同；
    接收所述策略决策实体发送的所述M个策略，从所述每个策略中获得所述每个策略的优化目标；
    根据所述第m个策略的优化目标和所述N个优化目标的优先级，确定所述第m个策略的优化目标的优先级。
  15. 根据权利要求14所述的策略执行实体，其特征在于，所述N个优化目标包括以下网络业务参数中的任意一个或其组合：网络业务NS实例的业务负载、NS实例的能耗、NS实例的虚拟网络功能转发图VNFFG的业务路由选择、NS实例的VNFFG的业务链路拥塞或NS实例的虚拟网络功能VNF故障。
  16. 根据权利要求10-15中任一项所述的策略执行实体，其特征在于，还包括：
    上报模块，用于向策略决策实体上报因执行冲突没有被执行的策略。
  17. 一种策略决策实体，其特征在于，包括：
    优先级制定模块，用于为M个策略的策略协调维度制定优先级，所述M个策略中每个策略包括：策略条件和执行动作，所述M个策略中的 第m个策略的策略协调维度包括以下信息中的至少一个：所述第m个策略、所述第m个策略的执行动作或者所述第m个策略的优化目标，所述第m个策略的优化目标为所述第m个策略的策略条件中的优化目标，M为大于或等于2的整数，m的取值为1，2，……，M；
    发送模块，用于将所述M个策略，以及所述M个策略的策略协调维度的优先级发送给策略执行实体。
  18. 根据权利要求17所述的策略决策实体，其特征在于，还包括：
    接收模块，用于接收所述策略执行实体上报的因执行冲突没有被执行的策略；
    更新模块，用于更新所述因冲突没有被执行的策略的策略协调维度的优先级。
  19. 一种策略执行实体，其特征在于，包括：处理器、存储器、通信接口和系统总线，所述存储器和所述通信接口通过所述系统总线与所述处理器连接并完成相互间的通信，所述存储器用于存储计算机执行指令，所述通信接口用于和其他设备进行通信，所述处理器用于运行所述计算机执行指令，使所述策略执行实体执行如权利要求1-7任一所述的方法。
  20. 一种策略决策实体，其特征在于，包括：处理器、存储器、通信接口和系统总线，所述存储器和所述通信接口通过所述系统总线与所述处理器连接并完成相互间的通信，所述存储器用于存储计算机执行指令，所述通信接口用于和其他设备进行通信，所述处理器用于运行所述计算机执行指令，使所述策略决策实体执行如权利要求8或9任一所述的方法。
PCT/CN2015/080972 2015-06-08 2015-06-08 Nfv系统中的策略协调方法和装置 WO2016197301A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2015/080972 WO2016197301A1 (zh) 2015-06-08 2015-06-08 Nfv系统中的策略协调方法和装置
CN201580042760.1A CN106576260B (zh) 2015-06-08 2015-06-08 Nfv系统中的策略协调方法和装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2015/080972 WO2016197301A1 (zh) 2015-06-08 2015-06-08 Nfv系统中的策略协调方法和装置

Publications (1)

Publication Number Publication Date
WO2016197301A1 true WO2016197301A1 (zh) 2016-12-15

Family

ID=57502815

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/080972 WO2016197301A1 (zh) 2015-06-08 2015-06-08 Nfv系统中的策略协调方法和装置

Country Status (2)

Country Link
CN (1) CN106576260B (zh)
WO (1) WO2016197301A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019206234A1 (zh) * 2018-04-28 2019-10-31 华为技术有限公司 Nfv策略协商方法及系统
WO2022089491A1 (en) * 2020-10-28 2022-05-05 Telefonaktiebolaget Lm Ericsson (Publ) Methods and apparatuses for instantiation of ns or vnf
US12034609B2 (en) 2020-10-28 2024-07-09 Telefonaktiebolaget Lm Ericsson Methods and apparatuses for instantiation of NS or VNF

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109391483A (zh) * 2017-08-04 2019-02-26 中国电信股份有限公司 网络实例化方法、网络功能虚拟化编排器和系统
CN110247783A (zh) * 2018-03-07 2019-09-17 中国移动通信有限公司研究院 一种扩缩容策略冲突处理方法及装置
CN108921369A (zh) * 2018-05-08 2018-11-30 阿里巴巴集团控股有限公司 冲突规则生成方法及装置和电子设备
CN111404704B (zh) * 2019-01-02 2023-05-09 中国移动通信有限公司研究院 Vnf扩缩容方法及装置、网元及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1391674A (zh) * 1999-10-20 2003-01-15 阿尔卡塔尔公司 用于数据通信交换机的交换机上政策规则高速缓存
CN103188168A (zh) * 2011-12-31 2013-07-03 中国移动通信集团公司 一种QoS优化方法和设备
CN103763197A (zh) * 2014-01-27 2014-04-30 杭州华三通信技术有限公司 一种流表项冲突检测装置和方法
CN104104615A (zh) * 2014-07-21 2014-10-15 华为技术有限公司 策略冲突解决方法以及装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9838483B2 (en) * 2013-11-21 2017-12-05 Oracle International Corporation Methods, systems, and computer readable media for a network function virtualization information concentrator

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1391674A (zh) * 1999-10-20 2003-01-15 阿尔卡塔尔公司 用于数据通信交换机的交换机上政策规则高速缓存
CN103188168A (zh) * 2011-12-31 2013-07-03 中国移动通信集团公司 一种QoS优化方法和设备
CN103763197A (zh) * 2014-01-27 2014-04-30 杭州华三通信技术有限公司 一种流表项冲突检测装置和方法
CN104104615A (zh) * 2014-07-21 2014-10-15 华为技术有限公司 策略冲突解决方法以及装置

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019206234A1 (zh) * 2018-04-28 2019-10-31 华为技术有限公司 Nfv策略协商方法及系统
WO2022089491A1 (en) * 2020-10-28 2022-05-05 Telefonaktiebolaget Lm Ericsson (Publ) Methods and apparatuses for instantiation of ns or vnf
US12034609B2 (en) 2020-10-28 2024-07-09 Telefonaktiebolaget Lm Ericsson Methods and apparatuses for instantiation of NS or VNF

Also Published As

Publication number Publication date
CN106576260A (zh) 2017-04-19
CN106576260B (zh) 2020-06-26

Similar Documents

Publication Publication Date Title
WO2016197301A1 (zh) Nfv系统中的策略协调方法和装置
WO2018213991A1 (zh) 网络切片创建的方法、装置以及通信系统
US10230765B2 (en) Techniques to deliver security and network policies to a virtual network function
US20170141944A1 (en) Verifier for network function virtualization resource allocation
JP6466003B2 (ja) Ｖｎｆフェイルオーバの方法及び装置
EP2957068B1 (en) Methods, systems, and computer readable media for providing a virtualized diameter network architecture and for routing traffic to dynamically instantiated diameter resource instances
Mann et al. Remedy: Network-aware steady state VM management for data centers
EP3800926B1 (en) Alarm method and device
US9588815B1 (en) Architecture for data collection and event management supporting automation in service provider cloud environments
US10057109B2 (en) Defining interdependent virtualized network functions for service level orchestration
EP3119034B1 (en) Fault handling method, device and system based on network function virtualization
US20190116110A1 (en) Location Based Test Agent Deployment In Virtual Processing Environments
WO2016165292A1 (zh) 一种实现虚拟网络功能部署规格配置的方法及装置
US11068298B2 (en) Hardware acceleration method and related device
CN108370328B (zh) 一种nfv mano策略描述符的管理方法及装置
US11909603B2 (en) Priority based resource management in a network functions virtualization (NFV) environment
JP2018528526A (ja) 複合サービスファンクションチェーン間の負荷及びソフトウェア構成の制御
CN104584484A (zh) 提供基于策略的数据中心网络自动化的系统和方法
CN109074280B (zh) 网络功能虚拟化
CN107251486A (zh) 一种扩展联动的方法、装置及系统
EP3534578B1 (en) Resource adjustment method, device and system
CN107534577B (zh) 一种网络业务实例化的方法及设备
CN104539551A (zh) 路由交换平台虚拟服务迁移方法
CN108965147B (zh) 网络控制方法、装置及网络设备
KR20220104241A (ko) 네트워크 작업 방법, 장치, 설비 및 저장매체

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: 15894571

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: 15894571

Country of ref document: EP

Kind code of ref document: A1