WO2015110083A1

WO2015110083A1 - Virtual network resource management

Info

Publication number: WO2015110083A1
Application number: PCT/CN2015/071593
Authority: WO
Inventors: Fengxian HUANG; Songer SUN
Original assignee: Hangzhou H3C Technologies Co., Ltd.
Priority date: 2014-01-27
Filing date: 2015-01-27
Publication date: 2015-07-30
Also published as: CN104811328B; CN104811328A

Abstract

The virtual network management of the present disclosure may include: for each virtual network function instance, the primary management center unit may assign resources nodes implementing the virtual network function instance into at least one region according to a predetermined strategy; select a region center node for each region according to a predetermined region center node selection principle; receive a resources adjustment request sent by a region center node, and adjust resources of a region corresponding to the region center node.

Description

VIRTUAL NETWORK RESOURCE MANAGEMENT

BACKGROUND

With the development of technologies， such as the virtualization， cloud computing and software defined network (SDN) ， the network virtualization become a trend. The implementation of the cloud computing is based on a large resources pool. To build a flexible， diverse， automatic， simple and valuable cloud system， all the computing devices， storing devices and network devices may be abstracted into resources.

For the above-mentioned reasons， the network functions virtualization (NFV) is provided to match the relationship between the resources of physical devices and network virtualization. NFV first decouples the physical device and the virtual resources， and then may achieve a variety of network functions through software running on standardized servers and switch devices. For example， the network functions may include virtual network address translation (vNAT) ， virtual firewall (vFW) ， and virtual router (vRouter) etc.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the present disclosure， reference should be made to the Detailed Description below， in conjunction with the following drawings in which like reference numerals refer to corresponding parts throughout the figures.

Figure 1 is a schematic diagram illustrating a system for managing virtual network resources according to an example of the present disclosure.

Figure 2a is a flow diagram illustrating a method for managing virtual network resources according to an example of the present disclosure.

Figure 2b is a flow diagram illustrating a method for managing virtual network resources according to another example of the present disclosure.

Figure 3 is a schematic diagram illustrating a region division according to an example of the present disclosure.

Figure 4 is a flow diagram illustrating a method for managing virtual network resources according to another more example of the present disclosure.

Figure 5a and figure 5b are schematic diagrams illustrating a device for managing virtual network resources according to an example of the present disclosure.

Figure 6a and figure 6b are schematic diagrams illustrating another device for managing virtual network resources according to an example of the present disclosure.

DETAILED DESCRIPTION

Reference will now be made in detail to examples， which are illustrated in the accompanying drawings. In the following detailed description， numerous specific details are set forth in order to provide a thorough understanding of the present disclosure. Also， the figures are illustrations of an example， in which modules or procedures shown in the figures are not necessarily essential for implementing the present disclosure. In other instances， well-known methods， procedures， components， and circuits have not been described in detail so as not to unnecessarily obscure aspects of the examples.

In an example of the present disclosure， in order to achieve dynamic and extensible virtual network resources management， when a virtual network function instance is added， the region division is performed and a region center node of resources nodes corresponding to the virtual network function instance is selected. Such that part of operations for managing the resources nodes in the region may be offloaded to the region center node. Based on the present disclosure， not only the dynamic and extensible virtual network resources management is achieved， but also the management burdens of the primary management center unit may be reduced.

Figure 1 is a schematic diagram illustrating a system for managing virtual network resources according to an example of the present disclosure. As shown in figure 1， the system may include a primary management center unit， a virtual network function unit and a virtual resources pool.

In figure 1， the virtual layer hides the differences of the hardware， such as physical devices by abstracting them as virtual resources. The cloud management platform is configured to define the cloud services. The primary management center unit is configured to manage each virtual network resource node in the virtual resources pool. The virtual network function unit comprises virtual network function instances and is configured to provide various network services. The virtual resources pool is configured to provide virtual network resources nodes which are used to implement a virtual network function instance.

The virtual network resources mentioned in the present disclosure refer to virtual resources formed from the resource virtualization of physical computing devices， network devices and storage devices. The virtual resources may be divided into multiple virtual units according to the user requirements. After the virtual units are pooled into a virtual resource pool， each virtual unit in the virtual resource pool may be called a resources node. The term “resources node” as used in the present disclosure， refers to a virtual network resources node and may include a virtual computing and input/output (IO) unit. The resources nodes may be created according to a minimization principle or user requirements. The resources nodes are thus virtual resources entities which are capable of being managed by a primary management center unit and a region center node. The resources node may for example be a virtual machine hosted on a server or a virtualized physical forwarding device capable of being managed by a virtual network controller. The described in the present disclosure may also be a virtual computing and IO node in the form of a virtual machine.

In the whole system， the resources management function of the primary center management unit is one aspect of the present disclosure. The primary center management unit is to support the distribution and state monitoring of the resources node， and also to manage the creation， monitoring and resources decision adjustment of all virtual network function instances， such as Vfw， vRouter， and so on.

Figure 2a is a flow diagram illustrating a method for managing virtual network resources according to an example of the present disclosure. The method may be applied to the system shown in figure 1. As shown in figure 2a， the method may include the following processes.

In block 101， a primary management center unit may assign resources nodes which implement a same virtual network function instance into at least one region according to a predetermined strategy. In an example， resources nodes which implement the same one virtual network function instance may be assigned into multiple regions.

In block 102， the primary management center unit may select a region center node for each region according to a predetermined region center node selection principle.

In block 103， the primary management center unit may receive a resources adjustment request sent by a region center node and adjust node resources of the corresponding region.

Figure 2b is a flow diagram illustrating a method for managing virtual network resources according to another example of the present disclosure. The method may be applied to the system shown in figure 1. As shown in figure 2a， the method may include the following processes.

In block 201， the primary management center unit may receive a registration request packet sent by a resources node after the resources node is started， the registration request packet carries information of the resources node； then the primary management center unit may record the information of the resources node and put the resources node into the virtual resources pool.

In practice， a resources node may be found by a network management system， when a resources node which may be added into the virtual resources pool is found， the resources node may be configured the address of the primary management center unit and may be indicated to send the information to be registered to the primary management center unit. Thus， when the resources node is started， the resources node may send the registration request to the primary management center unit.

In an example， the information of the resources node carried in the registration request includes but is not limited to the following information.

1) An Identifier (ID) of the resources node which is used to uniquely identify the resources node.

2) A Media Access Control (MAC) address of the resources node.

3) Central Processing Unit (CPU) capacity of the resources node， wherein the CPU capacity may include dominant frequency， the number of processor cores.

4) The memory capacity of the resources node.

In block 202， the primary management center unit may， according to service requirements of the user or cloud service requirements defined by the cloud management platform， create a virtual network function instance corresponding to the service requirements by utilizing multiple resources nodes in the virtual resources pool.

In block 203， for each virtual network function instance， the primary management center unit may assign the resources node implementing the virtual network function instance into at least one region according to a predetermined strategy.

The specific processing method in block 101 or block 203 includes but is not limited to the followings.

First: the primary management center unit may perform region division of the virtual network function instances according to the services provided by the virtual network function instances. The primary management center unit may determine whether a virtual network function instance provides multiple services and if the virtual network function instance provides multiple services， may assign resources nodes implementing different services of the virtual network function instance into different regions. The primary management center unit may determine whether a number of resources nodes implementing a same service ora same virtual network function instance is less than a preset region capacity reference value； assign said resources nodes implementing a same service of a same virtual network function instance to a single region if said number is below the preset region capacity reference value and assign said resources nodes implementing a same service of a same virtual network capacity reference value to a plurality of regions in response to determining that said number is above the preset region capacity reference value.

For example， the primary management center unit may first decide whether the virtual network function instance provides multiple services， when the virtual network function instance provides multiple services， assign the resources nodes implementing the same service into a same region， namely the number of the regions is the same as the number of the services； otherwise， assign all resources nodes implementing the virtual network function instance into one region.

Second: the primary management center unit may， according to the services provided by the virtual network function instance and the number of the resources nodes implementing the virtual network function instance， perform region division of the virtual network function instance.

In an example， when the virtual network function instance provides multiple services， the primary management center unit may assign the resources nodes implement the same service into a same region， namely， the number of the regions is the same as the number of the services； when the virtual network function instance provides one service and the number of resources nodes implementing the virtual network function instance is larger than a preset region capacity reference value， the primary management center unit may assign the resources nodes implementing the virtual network function instance into multiple regions according to the region capacity reference value； otherwise， assign all resources nodes implementing the virtual network function instance into one region.

In another example， when the virtual network function instance provides multiple services， the primary management center unit may first pre-assign resources nodes implementing the same service into one region， namely， the number of the regions is the same as the number of the services； and then check the number of resources nodes in each region， when the number of resources nodes in a region is larger than a preset region capacity reference value， the primary management center unit may further assign the resources nodes in the region into multiple regions according to the region capacity reference value； when the virtual network function instance provides only one service and the number of resources nodes implementing the virtual network function instance is larger than the preset region capacity reference value， the primary management center unit may assign the resources nodes implementing the virtual network function instance into multiple regions according to the region capacity reference value； otherwise， the primary management center unit may assign all resources nodes implementing the virtual network function instance into one region.

Figure 3 is a schematic diagram illustrating a region division according to an example of the present disclosure. As shown in figure 3， in this example， the virtual network function unit includes three virtual network function instances， namely vNAT， vFW and virtual Intrusion Prevention System (vIPS) . Assuming that the vNAT provides one service and the number of resources nodes implementing the vNAT is not larger than the preset region capacity reference value， so the resources nodes implementing the vNAT are assigned into one region， which is labeled as region 1. Assuming that the vFW provides two services， such as packet filtering and Deep Packet Inspection (DPI) ， the resources nodes implementing the vFW are assigned into two regions， which are labeled respectively as region 2 and region 3. Assuming that the vIPS provides a service， but the number of resources nodes implementing the vIPS is larger than the preset region capacity reference value， the resources nodes implementing the vIPS are assigned into two regions， which are labeled respectively as region 4 and region 5， according to the preset region capacity reference value.

In block 204， the primary management center unit may select a region center node for each region corresponding to each virtual network function instance according to a predetermined region center node selection principle.

The region center node selection principle mentioned in block 102 and block 204 includes but is not limited to the following: selecting the region center node based on the CPU capacity， memory capacity and/or MAC address of the resources nodes in the region.

For example， at first， the primary management center unit may compare the CPU capacities of resources nodes in a region and select a resource node of which the CPU capacity is the highest as the region center node. Specifically， the CPU dominant frequencies of resources nodes in the region may be compared， and the resource node of which the CPU dominant frequency is the highest may be selected as the region center node. When the CPU dominant frequencies of a plurality of resources nodes in the region are the same， then the number of processor cores of the plurality of resources nodes in the region may be compared， and the resource node of which the number of processor cores is the largest may be selected as the region center node.

When the CPU capacities of a plurality of resources nodes in the region are the same， the memory capacities of the plurality of resources nodes in the region may be compared， and a resources node of which a memory capacity is the largest may be selected as the region center node.

When the CPU capacities and the memory capacities of a plurality of resources nodes in the region are the same， the medium access control (MAC) addresses of the plurality of resources nodes in the region may be compared， and a resources node of which the MAC address is the highest or lowest may be selected as the region center node.

In practice， two region center nodes may be selected for each region and may be taken as the primary region center node and standby region center node respectively. At first， the primary region center node may be selected according to the region center node selection principle and then the standby region center node may be selected according to the region center node selection principle.

In block 205， the primary management center unit may send a region node list notification packet to the selected region center node. The region node list notification packet carries a list of resource nodes in the region.

When selecting two region center nodes acted as the primary region center node and the standby region center node for the region respectively， the primary management center unit may send the list of resource nodes in the region to the primary region center node. In the list of resource nodes carried in the packet， the standby region center node role of the standby region center node may be indicated.

In block 206， the primary management center unit may receive and record region state information reported by each region center node， and then generate a region display.

The region display refers to state information of each region is recorded and displayed according to network management requirements.

It should be noted that the primary management center unit may support the combination of state information of multiple regions corresponding to the same virtual network function instance， so as to conveniently monitor the running conditions of the same virtual network function instance.

In block 207， the primary management center unit may receive a resources adjustment request sent by a region center node and adjust resources nodes of a region where the region center node is located.

Specifically， in block 103 and block 207， the primary management center unit may receive a resources addition request packet sent by a region center node. And then according to a resources node identifier carried in the resources addition request packet， the primary management center unit may search the virtual resources pool for the corresponding resources node. When the resources node is found， according to the number of resources nodes required to be added carried in the resources addition request packet， the primary management center may obtain resources nodes of corresponding number from the virtual resources pool and allocate the obtained resources nodes to the region. The primary management center unit may receive a resources reduction request packet sent by a region center node. And then according to resources node identifiers and the number of resources nodes requested to be removed carried in the resources reduction request packet， the primary management center unit may remove resources nodes of corresponding number from the region and put the recycled resources nodes (namely the resources nodes removed from the region) into the virtual resources pool.

Figure 4 is a flow diagram illustrating a method for managing virtual network resources according to another more example of the present disclosure. The method may be applied to the system shown in figure 1. As shown in figure 4， the method may include the following processes.

In block 401， the region center node may receive a region node list notification packet sent by the primary management center unit， and send a region joining notification packet to each resources node according to a list of resource nodes carried in the region node list notification packet； and then receive a region joining acknowledgement packet returned by each resources node. The region joining acknowledgement packet carries information of the resources node.

The information of the resources node carried in the region joining acknowledgement packet may be consistent with that carried in the registration request packet sent by the resources node in block 201.

Ifthe primary management center unit selects two region center nodes for the region， namely the primary region center node and the standby region center node， the standby region center node role of the standby region center node may be indicated in the list of resource nodes carried in the region node list notification packet. Simultaneously， in block 401， after receiving the region node list notification packet sent by the primary management center unit， the primary region center node may indicate the standby region center node role of the standby region center node in the region joining notification packet sent to the standby region center node， and synchronize the list of resources nodes in the region on the standby region center node.

In block 402， the region center node may monitor the resources usage rate of the region. When the resources usage rate reaches to a service bearing capacity threshold of the region， the region center node may send a resources adjustment request to the primary management center unit.

Specifically， the region center node may make a resources adjustment decision in the region according to state information reported by resources nodes in the region. Then the region center node may send the primary management center unit a resources adjustment request packet， such as a resource addition request packet or a resources reduction request packet. The resources adjustment request packet carries identifiers and the number of resources nodes to be added or removed.

For example， the region center node may monitor the resources usage rate of the region according to state information reported by resources nodes in the region. When finding that the resources usage rate reaches to an upper limit of the service bearing capacity of the region， the region center node may send a resources addition request carrying identifiers and the number of resources nodes to the primary management center unit； when finding that the resources usage rate is less than an lower limit of the service bearing capacity of the region， the region center node may send a resources reduction request carrying identifiers and the number of resources nodes to the primary management center unit.

In addition， each resources node in each region may periodically report state information of the resources node to the region center node of the region. Each region center node may summarize the state information reported by all resources node in the region periodically or a preset condition is satisfied to make a resources adjustment decision in the region. Subsequently， the region center node may report summarized region state information to the primary management center unit.

The state information of the resources node includes: resources state， service performance state， log， warning and the like. The resources state may include: CPU capacity， memory capacity， etc. The service performance state may include the concurrent connection capacity， connection establishment rate and so on.

When the region includes two region center nodes， namely the primary region center node and the standby region center node. The primary region center node may real-timely or periodically back up the state information of each resources node in the region and summarized region state information to the standby region center node. Additionally， in this example， the primary region center node and the standby region center node may periodically exchange a keeping alive packet. When the keeping alive packet exchange is failure， the standby region center node may be upgraded to a primary region center node and then may send a primary-standby switch packet to each resources node in the region to notify that the standby region center node becomes the primary region center node. Subsequently， each resources node in the region may report the state information to the new primary region center node.

An example will be given hereinafter to illustrate an application for monitoring the virtual network resources load and performing region decision.

In block 01， each resources node in the region periodically reports the load of the resources node to a region center node which is located at the region.

The load may include concurrent connection capacity， connection establishment rate and so on.

In block 02， the region center node periodically summarizes the load of each resources node in the region and reports summarized region load to the primary management center unit.

In block 03， the region center node monitors whether the region load reaches to the upper limit of the service loading capacity of the region or is less than the lower limit of the service loading capacity of the region. When the region load reaches to the upper limit of the service loading capacity of the region， block 04 is performed； when the region load is less than the lower limit of the service loading capacity of the region， block 06 is performed.

In block 04， the region center node sends a resources addition request packet to the primary management center unit， wherein the resources addition request packet carries identifiers of resources node and the number of resources nodes required to be added.

In block 05， the primary management center unit receives the resources addition request packet sent by the region center node， and then searches the virtual resources pool for corresponding idle resources nodes according to the identifiers of resources nodes which are carried in the resources addition request packet. When finding the idle resources nodes， the primary management center unit selects the idle resources nodes from the virtual resources pool according to the number of resources nodes required to be added carried in the packet， and then allocates the selected idle resources nodes to the region.

Ifthere is no idle resources nodes found in the virtual resources pool， it means that the resources nodes have run out and then the primary management center unit may return a response packet to the region center node which carries information indicating that the resources nodes have run out. Ifthere are some idle resources nodes found in the virtual resources pool， but the number of corresponding idle resources nodes in the virtual resources pool is less than the number of resources nodes required to be added， thus the primary management center unit may allocate the corresponding idle resources nodes in the virtual resources pool to the region. And at the same time， the primary management center unit may return a response packet to the region center node which carries information indicating that there is no enough idle resources nodes in the virtual resources pool.

In block 06， the region center node sends a resources reduction request packet to the primary management center unit； wherein the resources reduction request packet carries identifiers of resources node and the number of resources nodes required to be removed.

In block 07， the primary management center unit receives the resources reduction request packet sent by the region center node， and then removes resources nodes of corresponding number from the region according to the identifiers and the number of resources nodes required to be removed carried in the resources reduction request packet， and puts removed resources nodes into the virtual resources pool.

It should be noted that， in an example of the present disclosure， the packets interacted between the primary management center unit and the region center node or between the region center node and resources nodes in the region， may be encapsulated in Ethernet II format. Except the layer 2 protocol header， format of other parts of the packet may be shown in table 1.

Table 1

The meaning of each field is as follows:

Version field means the version and may occupy 8 bits.

Code field means the type of the packet and may occupy 8 bits.

Length field means the length of the packet， namely a total length of Version， Code， Length， Identifier and Payload and may occupy 2 bytes.

Identifier field means the identifier of the packet. The same identifier should be used in the request packet and the response packet. When the packet is retransmitted， the identifier of the packet remains unchanged. The identifier field may occupy 8 bits.

Payload field means the specific contents of each type of packet and adopts the Type Length and Value (TLV) format， In which:

Type refers to the option type and may occupy 8 bits.

Length refers to the option length and may occupy 8 bits. The length is the total length of Type， Length and Value.

Value refers to the option value.

Corresponding to figures 2a and 2b， the present disclosure also provides a virtual network resources management logic. As shown in figure 5a， the virtual network resources management logic may run in a device which includes a memory 510 and a processor 520 in communication with the memory 510. The memory 510 stores the virtual network resources management logic.

By executing the computer executable instructions corresponding to the virtual network resources management logic， the processor 520 is to: for each virtual network function instance， assign resources nodes implementing the same virtual network function instance into at least one region according to a predetermined strategy； select a region center node for each region according to a predetermined region center node selection principle； and receive a resources adjustment request sent by a region center node and adjust resources nodes of a corresponding region where the region center node is located.

The processing for assigning resources nodes implementing the virtual network function instance into at least one region according to the predetermined strategy may include: the primary management center unit may determine whether a virtual network function instance provides multiple services and ifthe virtual network function instance provides multiple services， may assign resources nodes implementing different services of the virtual network function instance into different regions. The primary management center unit may determine whether a number of resources nodes implementing a same service of a same virtual network function instance is less than a preset region capacity reference value； assign said resources nodes implementing a same service of a same virtual network function instance to a single region if said number is below the preset region capacity reference value and assign said resources nodes implementing a same service of a same virtual network capacity reference value to a plurality of regions in response to determining that said number is above the preset region capacity reference value.

In an example， the processing for assigning resources nodes implementing the virtual network function instance into at least one region according to the predetermined strategy may include any one of the following processing methods.

The first method: a determination is made as to whether the virtual network function instance provides multiple services. When the virtual network function instance provides multiple services， resources nodes implementing each service may be assigned into one region， namely the number of the regions is the same as the number of the services. Otherwise， all resources nodes implementing the virtual network function instance may be assigned into one region.

The second method: when the virtual network function instance provides multiple services， resources nodes implementing each service may be assigned into one region， namely the number of the regions is the same as the number of the services. When the virtual network function instance provides one service and the number of resources nodes implementing the virtual network function instance is larger than a preset region capacity reference value， the resources nodes implementing the virtual network function instance may be assigned into multiple regions according to the region capacity reference value. Otherwise， all resources nodes implementing the virtual network function instance may be assigned into one region.

The third method: when the virtual network function instance provides multiple services， resources nodes implementing each service may be pre-assigned into one region， namely the number of the regions is the same as the number of the services， and then the number of resources nodes in each region is checked， when the number of resources nodes in a region is larger than a preset region capacity reference value， the resources nodes in the region may be assigned into multiple regions according to the region capacity reference value. When the virtual network function instance provides one service， and the number of resources nodes implementing the virtual network function instance is larger than a preset region capacity reference value， the resources nodes implementing the virtual network function instance may be assigned into multiple regions according to the region capacity reference value. Otherwise， all resources nodes implementing the virtual network function instance may be assigned into one region.

The region center node selection principle may include: selecting the region center node based on the CPU capacity， memory capacity and/or MAC address of the resources nodes in the region.

For example， the CPU capacities of resources nodes in a region may be compared， and a resource node of which the CPU capacity is the highest may be selected as the region center node.

When CPU capacities of a plurality of resources nodes in the region are the same， the memory capacities of the plurality of resources nodes in the region may be compared， and the resources node of which a memory capacity is the largest may be selected as the region center node.

When CPU capacities and memory capacities of a plurality of resources nodes in the region are the same， the MAC addresses of the plurality of resources nodes in the region may be compared， and the resources node of which the MAC address is the highest or lowest may be selected as the region center node.

After selecting a region center node for each region of the virtual network function instance， the processor 520 is further to: send a region node list notification packet to the selected region center node. The region node list notification packet carries a list of resource nodes in the region so that the region center node may notify the resources nodes in the region that they are assigned to said region.

The processing for adjusting resources nodes of a corresponding region where the region center node is located may include: when a resources addition request sent by a region center node is received， upon the idle resources nodes found in the virtual resources pool， the idle resources nodes are allocated to the region； when a resources reduction request sent by a region center node is received， upon the corresponding resources nodes removed from the region， thus the removed resources nodes are put into the virtual resources pool.

As shown in figure 5b， in an example， the virtual network resources management logic may include an instance creation module 501， a region division module 502， a region selection module 503， a region display module 504 and a region monitor module 505. Specifically， the processor 520 may achieve the following operations by executing the corresponding modules of the virtual network resources management logic.

By executing the instance creation module 501， the processor 520 is to: create a virtual network function instance corresponding to the service requirements utilizing multiple resources nodes in the virtual resources pool according to service requirements of a user or cloud service requirements defined by the cloud management platform.

By executing the region division module 502， the processor 520 is to: assign the multiple resources nodes implementing the same virtual network function instance which is created by executing the instance creation module 501 into at least one region according to a predetermined strategy. In an example， resources nodes which implemented one virtual network function instance may be assigned into multiple regions.

By executing the region selection module 503， the processor 520 is to: select a region center node for the region according to a predetermined region center node selection principle.

By executing the region display module 504， the processor 520 is to: receive region state information reported by each region center node selected by executing the region selection module 503 and generate a region display.

By executing the region monitor module 505， the processor 520 is to: receive a resources adjustment request sent by a region center node selected by executing the region selection module 503 and adjust resources of corresponding region. For instance， adjust resources nodes of a region where the region center node is located.

When the region division module 502 is executed to assign the multiple resources nodes implementing the same virtual network function instance into at least one region according to a predetermined strategy， The region division module 502 may determine whether a virtual network function instance provides multiple services and ifthe virtual network function instance provides multiple services， assign resources nodes implementing different services of the virtual network function instance into different regions. The region division module 502 may determine whether a number of resources nodes implementing a same service of a same virtual network function instance is less than a preset region capacity reference value； assign said resources nodes implementing a same service of a same virtual network function instance to a single region if said number is below the preset region capacity reference value and assign said resources nodes implementing a same service of a same virtual network capacity reference value to a plurality of regions in response to determining that said number is above the preset region capacity reference value. In an example， the region division module 502 may determine whether the virtual network function instance provides multiple services. When the virtual network function instance provides multiple services， the region division module 502 may assign resources nodes implementing each service into one region， namely the number of the regions is the same as the number of the services. Otherwise， the region division module 502 may assign all resources nodes implementing the virtual network function instance into one region.

In another example， the region division module 502 may be used to: when the virtual network function instance provides multiple services， assign resources nodes implementing each service into one region， namely the number of the regions is the same as the number of the services； when the virtual network function instance provides one service and the number of resources nodes implementing the virtual network function instance is larger than the preset region capacity reference value， assign the resources nodes implementing the virtual network function instance into multiple regions according to the region capacity reference value. Otherwise， the region division module 502 is used to assign all resources nodes implementing the virtual network function instance into one region.

In another more example， the region division module 502 may be used to: when the virtual network function instance provides multiple services， pre-assign resources nodes implementing each service into one region， namely the number of the regions is the same as the number of the services； and then check the number of resources nodes of each region， when the number of resources nodes of a region is larger than the preset region capacity reference value， assign the resources nodes in the region into multiple regions according to the region capacity reference value； when the virtual network function instance provides one service and the number of resources nodes implementing the virtual network function instance is larger than the preset region capacity reference value， assign the resources nodes implementing the virtual network function instance into multiple regions according to the region capacity reference value. Otherwise， the region division module 502 may be used to assign all resources nodes implementing the virtual network function instance into one region.

The region center node selection principle may include: selecting the region center node based on the CPU capacity， memory capacity and/or MAC address of the resources nodes in the region. For example， the CPU capacities of resources nodes in a region may be compared， and the resource node of which the CPU capacity is the highest may be selected as the region center node.

When executing the region selection module 503 to select a region center node for each region， the region selection module 503 may be used to select a primary region center node and a standby region center node for each region regarding to each virtual network function instance. In an example， one primary region center node and at least one standby region center node may be selected for each region regarding to each virtual network function instance.

After selecting a region center node for each region by executing the region selection module 503， the region selection module 503 may further be used to send a region node list notification packet to the selected region center node. The region node list notification packet carries a list of resource nodes in the region， so that the region center node may notify the resources nodes in the region that they are assigned to said region.

When executing the region selection module 503 to select the primary region center node and the standby region center node for each region， the standby region center node role of the standby region center node may be indicated in the list of resource nodes carried in the packet.

When executing the region monitor module 505 to adjust resources of corresponding region according to a resources adjustment request sent by a region center node， the region monitor module 505 may be used to， after receiving a resources addition request sent by a region center node， allocate the idle resources nodes to the region upon finding the corresponding idle resources nodes in the virtual resources pool. After receiving a resources reduction request sent by a region center node， the region monitor module 505 may be used to remove the corresponding resources nodes from the region and then put the removed resources nodes into the virtual resources pool.

When executing the instance creation module 501， the instance creation module 501 may further be used to receive a registration request packet sent by a resources node and then put the resources node into the virtual resources pool， wherein the registration request packet carries information of the resources node.

Corresponding to figure 4， the present disclosure also provides another virtual network resources management logic. As shown in figure 6a， the virtual network resources management logic may run in a device where a resources node is located， and forms a virtual network resources management device. The hardware architecture of the device may include a memory 610 and a processor 620 in communication with the memory 610. The memory 610 stores the virtual network resources management logic.

By executing the computer executable instructions corresponding to the virtual network resources management logic， the processor 620 is to receive a region node list notification packet sent by the primary management center unit and learn the resources node where the device is located according to the region node list notification packet； and according to a list of resource nodes carried in the region node list notification packet， send a region joining notification packet to each resources node in the list； receive a region joining acknowledgement packet returned by each resources node； wherein the region joining acknowledgement packet carries information of the resources node； monitor the resources usage rate of the region and send a resources adjustment request to the primary management center unit upon the resources usage rate reaching to a service bearing capacity threshold of the region.

Specifically， the processor 620 may be to monitor the resources usage rate of the region； when the resources usage rate reaches to an upper limit of the service bearing capacity of the region， send a resources addition request to the primary management center unit； when the resources usage rate is less than an lower limit of the service bearing capacity of the region， send a resources reduction request to the primary management center unit.

After receiving the region joining acknowledgement packet returned by each resources node， the processor may be further to periodically receive state information of a resources node reported by the resource node in the region and report summarized region state information of all resources nodes to the primary management center unit.

As shown in figure 6b， in an example， the virtual network resources management logic in the example may include a packet processing module 601 and a node monitoring module 602. Specifically， the processor 620 may achieve the following operations by executing the corresponding modules of the virtual network resources management logic running the device of which the resource nodes are located.

By executing the packet processing module 601， the processor 620 is to receive a region node list notification packet sent by the primary management center unit， the region node list notification packet carries a list of resource nodes； learn the resources node where the device is located is a region center node according to the region node list notification packet； and according to the list of resource nodes， send a region joining notification packet to each resources node in the list； receive a region joining acknowledgement packet returned by each resources node， the region joining acknowledgement packet carries information of the resources node； determine that the resources node joins to the region.

By executing the node monitoring module 602， the processor is to: monitor the resources usage rate of the region， and when the resources usage rate reaches to a service bearing capacity threshold of the region， send a resources adjustment request to the primary management center unit. Specifically， when finding that the resources usage rate reaches to an upper limit of the service bearing capacity of the region， send a resources addition request to the primary management center unit； when finding that the resources usage rate is less than an lower limit of the service bearing capacity of the region， send a resources reduction request to the primary management center unit.

By executing the node monitoring module 602， the processor is further to: receive state information of each resources node reported by the resource node joined in the region by executing the packet processing module 601， and report summarized region state information of all resources nodes to the primary management center unit.

When the region node list notification packet received by executing the packet processing module 601 indicates that a resources node is a standby region center node， sending the region joining notification packet to each resources node in the list by executing the packet processing module 601 may include: the standby region center node role of the standby region center node may be indicated in the region joining notification packet sent to each standby region center node.

After sending the region joining notification packet to each resources node in the list， by executing the packet processing module 601， the processor 620 may be further to: synchronize the list of resources nodes in the region on each standby region center node； periodically exchange a keeping alive packet with each standby region center node. When the keeping alive packet exchange is failure， one of the standby region center nodes may be upgraded to a primary region center node， and then send a primary-standby switch packet to each resources node in the region.

The methods and devices provided by examples of the present disclosure may be achieved by hardware， or machine-readable instructions， or a combination of hardware and machine-readable instructions. The machine-readable instructions may be stored， by at least one processor， in a non-transitory computer readable storage medium， such as a hard disk， a floppy disk， a magnetic disk， compact disk (e.g.， CD-ROM， CD-R， CD-RW， DVD-ROM， DVD-RAM， DVD-RW and DVD+RW) ， a tape， a nonvolatile memory card， ROM and other suitable storage device. Or part of the machine-readable instructions may be achieved by specific hardware， such as custom integrated circuits， gate arrays， FPGA， PLD， a computer of specific functions， etc.

Examples of the present disclosure also provides a computer readable storage medium， which is configured to store machine-readable instructions for making a computing device (may be a personal computer， a server， a network device， etc.， ) execute the method of the examples. Specifically， the system and device provided by the examples of the present disclosure all include a computer readable storage medium， which stores machine-readable program codes for achieving functions of above mentioned examples. The system and device (or CPU or MPU) may read out and execute the machine-readable program codes stored in the computer readable storage medium.

In this case， the program codes read from the computer readable storage medium can achieve anyone of above mentioned examples. Therefore， the program codes and the storage medium storing the program codes are a part of technical solutions of the present disclosure.

The storage medium for providing program codes includes a hard disk， a floppy disk， a magnetic disk， compact disk (e.g.， CD-ROM， CD-R， CD-RW， DVD-ROM， DVD-RAM， DVD-RW and DVD+RW) ， a tape， a Flash card， ROM， and so on. Optionally， it is possible to download the program codes from a server computer via a communication network.

It should be noted that， for the program codes being executed by a computer， a part of or all of operations achieved by the program codes may be implemented by an operating system running on a computer， so as to achieve technical solutions of anyone of example mentioned above. The computer executes instructions based on the program codes.

The program codes in the storage medium are written in a memory， and the memory is located in an extended board inside a computer， or in an extended unit connected with a computer. In the example， a CPU on the extended board or the extended unit executes a part of or all of the actual operations based on the program codes， so as to realize technical solutions of any of the aforementioned embodiments.

The foregoing description， for purpose of explanation， has been described with reference to specific examples. However， the illustrative discussions above are not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed. Many modifications and variations are possible in view of the above teachings. The examples were chosen and described in order to best explain the principles of the present disclosure and its practical applications， to thereby enable others skilled in the art to best utilize the present disclosure and various examples with various modifications as are suited to the particular use contemplated.

Claims

A method for managing virtual network resources， comprising：

for each virtual network function instance， a primary management center unit assigning resources nodes implementing the same virtual network function instance into at least one region according to a predetermined strategy；

，the primary management center unit selecting a region center node for each region according to a predetermined region center node selection principle； and

the primary management center unit receiving a resources adjustment request sent by a region center node and adjusting resources of a region where the region center node is located.
The method according to claim 1， comprising：

the primary management center unit determining whether a virtual network function instance provides multiple services and if the virtual network function instance provides multiple services， assigning resources nodes implementing different services of the virtual network function instance into different regions.
The method according to claim 1， comprising the primary management center unit determining whether a number of resources nodes implementing a same service of a same virtual network function instance is less than a preset region capacity reference value； assigning said resources nodes implementing a same service of a same virtual network function instance to a single region if said number is below the preset region capacity reference value and assigning said resources nodes implementing a same service of a same virtual network capacity reference value to a plurality of regions in response to determining that said number is above the preset region capacity reference value.
The method according to claim 1， wherein the region center node selection principle comprises：

selecting the region center node based on the CPU capacity， memory capacity and/or MAC address of the resources nodes in the region.
The method according to claim 1， after the primary management center unit selecting a region center node for each region，：

the primary management center unit sending a region node list notification packet to the region center node； the region node list notification packet carrying a list of resource nodes in the region， so that the region center node may notify the resources nodes in the region that they are assigned to said region.
The method according to claim 1， wherein the primary management center unit receiving a resources adjustment request sent by a region center node， and adjusting resources of a region where the region center node is located， comprises：

the primary management center unit receiving a resources addition request sent by a region center node and allocating idle resources nodes from a virtual resource pool to the region；

the primary mamagement center unit receiving a resources reduction request sent by a region center node， removing corresponding resources nodes from the region， and putting the removed resources nodes into the virtual resources pool.
The method according to claim 1， before the primary management center unit dividing resources nodes implementing the same virtual network function instance into a region according to a predetermined strategy；

the primary management center unit receiving a registration request packet sent by a resources node， the registration request packet carrying information of the resources node；

the primary management center unit putting the resources node into the virtual resources pool； and

the primary management center unit creating a virtual network function instance utilizing resources nodes in the virtual resources pool.
A device for managing virtual network resources， comprising： a processor and a non-transitory machine readable storage medium storing instructions executable by the processor to：

for each virtual network function instance， assign resources nodes implementing a same virtual network function instance into at least one region according to a predetermined strategy；

select a region center node for each region according to a predetermined region center node selection principle； and

receive a resources adjustment request sent by a region center node， and adjust resources of a region where the region center node is located.
The device according to claim 8， wherein the instructions further comprise instructions to determine whether a virtual network function instance provides multiple services and if the virtual network function instance provides multiple services， assign resources nodes implementing different services of the virtual network function instance into different regions.
The device according to claim 8， wherein the instructions further comprise instructions to determine whether a number of resources nodes implementing a same service of a same virtual network function instance is less than a preset region capacity reference value； assign said resources nodes implementing a same service of a same virtual network function instance to a single region if said number is below the preset region capacity reference value and assign said resources nodes implementing a same service of a same virtual network capacity reference value to a plurality of regions in response to determining that said number is above the preset region capacity reference value.
The device according to claim 8， wherein the region center node selection principle comprises：

selecting the region center node based on the CPU capacity， memory capacity and/or MAC address of the resources nodes in the region.
The device according to claim 8， after selecting a region center node for each region，

send a region node list notification packet to the region center node； the region node list notification packet carries a list of resource nodes in the region， so that the region center node may notify the resources nodes in the region that they are assigned to said region.
The device according to claim 8， wherein receive a resources adjustment request sent by a region center node， and adjust resources of a region where the region center node is located， comprises：

receive a resources addition request sent by a region center node， when corresponding idle resources nodes are found in the virtual resources pool， allocate the idle resources nodes to the region；

receive a resources reduction request sent by a region center node， remove corresponding resources nodes from the region， and put the removed resources nodes into the virtual resources pool.
A device for managing virtual network resources， comprising： a processor and a non-transitory machine readable storage medium storing instructions executable by the processor to：

receive a region node list notification packet sent by a primary management center unit， learn the resources node where the device is located is a region center node according to the region node list notification packet； and according to a list of resource nodes carried in the region node list notification packet， send a region joining notification packet to each resources node in the list； receive a region joining acknowledgement packet returned by each resources node； the region joining acknowledgement packet carries information of the resources node； and

monitor the resources usage rate of the region， and when the resources usage rate reaches to a service bearing capacity threshold of the region， send a resources adjustment request to the primary management center unit.
The device according to claim 14， wherein monitor the resources usage rate of the region， and when the resources usage rate reaches to a service bearing capacity threshold of the region， send a resources adjustment request to the primary management center unit， comprises：

monitor the resources usage rate of the region；

when the resources usage rate reaches to an upper limit of the service bearing capacity of the region， send a resources addition request to the primary management center unit； and

when the resources usage rate is less than an lower limit of the service bearing capacity of the region， send a resources reduction request to the primary management center unit.