WO2017181936A1 - 网络资源部署的方法、装置、控制器及核心网设备 - Google Patents

网络资源部署的方法、装置、控制器及核心网设备 Download PDF

Info

Publication number
WO2017181936A1
WO2017181936A1 PCT/CN2017/080907 CN2017080907W WO2017181936A1 WO 2017181936 A1 WO2017181936 A1 WO 2017181936A1 CN 2017080907 W CN2017080907 W CN 2017080907W WO 2017181936 A1 WO2017181936 A1 WO 2017181936A1
Authority
WO
WIPO (PCT)
Prior art keywords
virtual machine
network
sriov
moved
resources
Prior art date
Application number
PCT/CN2017/080907
Other languages
English (en)
French (fr)
Inventor
常亮
Original Assignee
中兴通讯股份有限公司
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 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2017181936A1 publication Critical patent/WO2017181936A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/12—Avoiding congestion; Recovering from congestion
    • H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5061—Partitioning or combining of resources
    • G06F9/5077—Logical partitioning of resources; Management or configuration of virtualized resources
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system
    • G06F9/5088—Techniques for rebalancing the load in a distributed system involving task migration
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/131—Protocols for games, networked simulations or virtual reality

Definitions

  • the present disclosure relates to the field of network virtualization technologies, for example, to a method, an apparatus, a controller, and a core network device for network resource deployment.
  • Open Virtualized Switch is a common virtualization technology for network resources. It can comply with the OpenFlow protocol of the open source stream. Through virtualization, the virtualized management layer virtualizes the hardware network resources so that the application layer can be used uniformly. The overhead of computing and deploying network resources is mainly reflected in the virtualization management layer.
  • Single Root I/O Virtualization can be used to solve the problem of virtual sharing, that is, multiple virtual machines can share the same Peripheral Component Interconnect (PCI) hardware at the same time, allowing applications to be applied.
  • the virtual function (VF) of the layer directly uses the corresponding physical function (PF) in the PCI hardware to improve efficiency.
  • the calculation and deployment overhead of the network resources is mainly on the hardware device, and it needs the hardware network card of the SRIOV to support the door.
  • SRIOV can include PCI passthrough mode and Macvtap Passthrough mode.
  • Macvtap Passthrough mode needs to be forwarded through the intermediate analog layer, the efficiency is not high in PCI Passthrough mode, but it can support virtual machine hot migration, while PCI Passthrough mode does not support virtual machine hot migration.
  • SRIOV NICs In terms of application performance, although the performance of SRIOV NICs is better than that of OVS NICs, SRIOV NICs require specific hardware resources, which are expensive, and have low cost performance for a good network environment. Therefore, for a virtualized environment, the SRIOV NIC should be deployed on a virtual machine with high network load in the environment. However, the current network planning is to statically deploy the SRIOV NIC to a specific virtual machine, and the SRIOV NIC cannot be deployed reasonably dynamically. .
  • the present disclosure provides a method, an apparatus, a controller, and a core network device for deploying a network resource, which can solve the technical problem that the SRIOV network card cannot be deployed reasonably dynamically.
  • the embodiment provides a method for deploying network resources, which may include: collecting real-time data of all virtual machines in the network after deploying a single input and output virtualized SRIOV network card resource in the network resource pool to a specified network virtual machine.
  • the service status data is used to evaluate the current service running status of each virtual machine according to the real-time service status data of each virtual machine collected in the first preset period.
  • the virtual machine that has been deployed with the SRIOV NIC resource is used as the to-be-migrated virtual machine, and the current service running status of the virtual machine to be moved into the virtual machine is determined to be inferior to the The current service running status of the virtual machine to be moved out, and the number of the virtual machines to be moved into is less than or equal to the number of virtual machines to be moved out.
  • the SRIOV network card resources to be moved out of the virtual machine are migrated to the virtual machine to be moved into the virtual machine.
  • the step of deploying the single input and output virtualized SIOOV NIC resources in the network resource pool to the designated network virtual machine may include: deploying the SRIOV NIC resources in the network resource pool to a pre-stated load A virtual machine at the load threshold.
  • the method may further include: separately deploying open source for all the virtual machines in the network. Virtual exchange OVS NIC resources.
  • the method may further include: the SRIOV network card resource to be moved out of the virtual machine and the to-be-moved before the migrating the SIOOV NIC resource to be moved out to the virtual machine After the service data of the OVS network card resource of the virtual machine is synchronized, the SRIOV network card resource of the virtual machine to be moved out is deactivated.
  • the method may further include: synchronizing the service data of the OVS NIC resource to be moved into the virtual machine to The SRIOV NIC resource to be moved into the virtual machine is activated, and the SRIOV NIC resource to be moved into the virtual machine is activated.
  • the step of deploying the single root input and output virtualized SIOOV NIC resources in the network resource pool to the designated network virtual machine may include: deploying the SIOOV NIC resources in the network resource pool according to the PCI passthrough mode to the designated Network virtual machine.
  • the method may further include: switching the SRIOV NIC resource to be moved out of the virtual machine from the PCI pass-through mode, before the migrating the SIOOV NIC resource to be moved out of the virtual machine to the virtual machine. Go to Macvtap pass-through mode.
  • the method may further include: the SRIOV NIC resource to be moved into the virtual machine from Macvtap pass-through mode switches to PCI pass-through mode.
  • the step of collecting real-time service state data of all virtual machines in the network includes:
  • the real-time transmission protocol RTP packet number, the number of lost packets, and the jitter duration of all the virtual machines in the network are collected according to the second preset period, where the second preset period is smaller than the first preset period.
  • the step of evaluating the current service running status of each virtual machine according to the real-time service status data of each virtual machine collected in the first preset period may include: acquiring the current status of each virtual machine by using the following formula: Business Health Evaluation Parameter Si:
  • Traf(i), Lost(i), and Jett(i) are respectively the number of RTP packets, the number of lost packets, and the jitter duration of the virtual machine i collected in the first preset period
  • Traf ( Avg), Lost (avg), and Jett (avg) are the average of the number of RTP packets of all the virtual machines collected in the first preset period, the average of the number of lost packets, and the average of the jitter duration.
  • the determining, according to the current service running status of each virtual machine, the step of determining to be moved into the virtual machine may include: determining, according to the current service running condition evaluation parameter Si of each virtual machine, the virtual machine to be moved, wherein the to-be-moved The Si entering the virtual machine is greater than or equal to the Si of the virtual machine to be moved out.
  • the determining, according to the current service running status of each virtual machine in the network, the step of determining to be moved into the virtual machine may include: determining, according to the current service running status of each virtual machine in the network, The VMs to be moved into the virtual machine are equal to the number of the virtual machines to be moved out, and each of the VMs to be moved out corresponds to one to be moved into the virtual machine.
  • the embodiment further provides an apparatus for deploying network resources, which may include: an acquisition module, an acquisition module, and a migration module.
  • the collection module is configured to collect real-time service state data of all virtual machines in the network after the single-input virtualized SRIOV NIC resource in the network resource pool is deployed to the specified network virtual machine, and according to the first preset
  • the real-time business status data of each virtual machine collected during the period evaluates the current business running status of each virtual machine.
  • the obtaining module is configured to use the virtual machine that has deployed the SRIOV NIC resource as the to-be-migrated virtual machine, and determine the current service to be moved into the virtual machine according to the current service running status of each virtual machine, where the current service to be moved into the virtual machine is obtained.
  • the running status is inferior to the current running status of the virtual machine to be moved out, and the number of the virtual machines to be moved into is less than or equal to the number of virtual machines to be moved out.
  • the migration module is configured to migrate the SRIOV network card resources of the virtual machine to be moved out to the virtual machine to be moved into the virtual machine.
  • the apparatus may further include: a deployment module, configured to separately deploy open source virtual switching OVS network card resources for all virtual machines of the network.
  • a deployment module configured to separately deploy open source virtual switching OVS network card resources for all virtual machines of the network.
  • Deactivating a module configured to: before the collecting module migrates the SIOOV NIC resource of the virtual machine to be moved to the virtual machine to be migrated, the SRIOV NIC resource of the virtual machine to be moved out and the to-be-moved After the service data of the OVS network card resource of the virtual machine is synchronized, the SRIOV network card resource of the virtual machine to be moved out is deactivated.
  • the activation module is configured to synchronize the service data of the OVS network card resource to be moved into the virtual machine to the mobile device after the migration module migrates the SIOOV network card resources of the virtual machine to be migrated into the virtual machine.
  • the SRIOV NIC resources that are moved into the virtual machine are described, and the SRIOV NIC resources to be moved into the virtual machine are activated.
  • the collecting module may include: a deployment unit, configured to deploy the SRIOV NIC resources in the network resource pool to the network virtual machine according to the PCI passthrough mode.
  • the device may further include: a first switching module, configured to be in the migration module Before the SRIOV NIC resources of the virtual machine to be migrated are migrated to the virtual machine, the SRIOV NIC resources of the virtual machine to be moved out are switched from the PCI pass mode to the Macvtap pass mode.
  • a first switching module configured to be in the migration module Before the SRIOV NIC resources of the virtual machine to be migrated are migrated to the virtual machine, the SRIOV NIC resources of the virtual machine to be moved out are switched from the PCI pass mode to the Macvtap pass mode.
  • the embodiment further provides a controller, comprising: the device for network resource deployment according to any one of the preceding claims.
  • the embodiment further provides a core network device, including: the controller as above.
  • the embodiment further provides a computer readable storage medium storing computer executable instructions for performing any of the above methods for network resource deployment.
  • the embodiment further provides a core network device, the core network device including one or more processors, a memory, and one or more programs, the one or more programs being stored in the memory when being processed by one or more When the device is executed, execute any of the above methods of network resource deployment.
  • the core network device including one or more processors, a memory, and one or more programs, the one or more programs being stored in the memory when being processed by one or more When the device is executed, execute any of the above methods of network resource deployment.
  • the embodiment further provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions, when the program instructions are executed by a computer A method of causing the computer to perform any of the above described network resource deployments.
  • the method for deploying network resources of the present disclosure dynamically and reasonably deploys SRIOV NIC resources in a virtual machine according to the real-time service status of each virtual machine in the network, can achieve load balancing of all virtual machines in the network, and improve utilization of SRIOV NIC resources. effectiveness. Therefore, the technical problem that the SRIOV network card cannot be reasonably and dynamically deployed can be solved.
  • FIG. 1 is a flowchart of a method for deploying network resources according to an embodiment of the present invention.
  • FIG. 2 is a flow chart of an application example of a method for deploying network resources according to the embodiment.
  • FIG. 3 is a schematic diagram of a network structure corresponding to initial deployment of network card resources in the method for deploying network resources according to the embodiment.
  • FIG. 4 is a schematic diagram of a network structure corresponding to a network card resource mode switching according to a method for deploying network resources in the embodiment.
  • FIG. 5 is a schematic diagram of a network structure corresponding to network card resource migration in the method for deploying network resources according to the embodiment.
  • FIG. 6 is a schematic diagram of a network structure corresponding to the method for synchronizing network card resources according to the method for deploying network resources in the embodiment.
  • FIG. 7 is a schematic structural diagram of a network resource deployment method corresponding to the network card resource switching in the embodiment.
  • FIG. 8 is a schematic structural diagram of an apparatus for deploying network resources according to an embodiment of the present invention.
  • FIG. 9 is a schematic diagram of a general hardware structure of a core network device according to an embodiment of the present invention.
  • the method for deploying network resources in this embodiment may include: Step 110 - Step 130.
  • step 110 after the SRIOV NIC resources in the network resource pool are deployed to the specified network virtual machine, the real-time service state data of all the virtual machines in the network is collected, and the real-time collected according to the first preset period is collected.
  • the business status data evaluates the current business health of each virtual machine.
  • the current service running status of each virtual machine can be effectively evaluated, so that the SRIOV network card resources can be adjusted reasonably and dynamically according to the service running status of the virtual machine.
  • the current service running status of the virtual machine can represent the current load status of the virtual machine. Generally, the higher the current load of the virtual machine, the worse the current business running status.
  • the current service running status of the virtual machine may also represent parameters such as the current network quality of the virtual machine or the delay or packet loss rate of the virtual machine sending and receiving data.
  • the virtual machine that has deployed the SRIOV NIC resource is used as the virtual machine to be moved out, and the virtual machine to be moved into the virtual machine is determined in the virtual machine according to the current service running status of each virtual machine.
  • the current service running status of the virtual machine is inferior to or equal to the current service running status of the virtual machine to be moved out.
  • the virtual machine Based on the current business running status of the virtual machine, the virtual machine is selected to be moved into the virtual machine. Dynamically adjust SRIOV NIC resources.
  • the current service running state of all the virtual machines in the network is compared with the current service running state of the virtual machine to be migrated from which the SRIOV network card resources have been deployed, and the virtual to-be-moved virtuality that has been deployed with the SRIOV NIC resources is determined.
  • the virtual machine whose operating status is poor or equal, and the virtual machine is to be moved to the virtual machine, that is, the virtual machine to be moved into the virtual machine may be a virtual machine that does not deploy SRIOV NIC resources, or may be deployed with a SRIOV network card.
  • the virtual machine of the resource may be a virtual machine that does not deploy SRIOV NIC resources, or may be deployed with a SRIOV network card.
  • the current service running state of each virtual machine is compared in all the virtual machines of the network, and the virtual machine with the worst running state of the current service is most to be moved into the virtual machine, that is, the current service running to be moved into the virtual machine.
  • the situation is inferior to the current business health of other virtual machines in the network except the one to be moved into the virtual machine.
  • the SIOOV NIC resources of the virtual machine to be moved out are migrated to the virtual machine to be moved into the virtual machine. .
  • the network resource deployment method in this embodiment dynamically and reasonably deploys the SRIOV network card resources according to the real-time service status of the virtual machine, so as to achieve load balancing of all virtual machines in the network and improve the utilization efficiency of the SRIOV network card resources. It can solve the technical problem that the SRIOV network card cannot be deployed reasonably dynamically.
  • the step of deploying a single input and output virtualized SIOOV NIC resource in the network resource pool to the network virtual machine may include: deploying the SRIOV NIC resource in the network resource pool to pre-statistically The high load of the virtual machine.
  • the load of all the virtual machines in the network resource pool is calculated in advance, and the virtual machine whose load is higher than the load threshold is selected, and the SRIOV network card resources are deployed to the selected virtual machine.
  • the load threshold may be determined by the staff according to the experience value, or may be calculated by the system according to the historical load data of all the virtual machines in the network resource pool.
  • SRIOV NIC resources can be deployed on virtual machines with historical load higher than the load threshold, thereby optimizing initial deployment and reducing the number of subsequent SRIOV NIC resource dynamic adjustments to save dynamic tuning. The whole resource.
  • the method may further include: all virtual machines in the network Deploy OVS NIC resources separately.
  • OVS NIC resources separately for all virtual machines in the network.
  • SRIOV NIC resources since the SRIOV NIC is enabled over the OVS NIC, the SRIOV NIC is enabled on the virtual machine.
  • the capacity of the OVS NIC is not limited by hardware, so configure one is not enabled.
  • the OVS NIC does not affect the performance and services of the virtual machine.
  • OVS common NIC For a virtual machine that only deploys OVS NIC resources, enable the OVS common NIC to work.
  • the method may further include: after the SRIOV network card resource of the virtual machine to be moved out is synchronized with the service data of the OVS network card resource of the virtual machine to be moved out, the virtual machine to be moved out
  • the SRIOV NIC resources are deactivated.
  • the SRIOV NIC resource migration is required. Before the migration, you need to ensure that the service data of the SRIOV NIC and the OVS NIC on the VM are synchronized, that is, the configured Internet Protocol (IP) routing information is the same, and then the SRIOV NIC that is to be moved out of the VM is processed. The resources are deactivated, so that the OVS network card to be moved out of the virtual machine takes over the service.
  • IP Internet Protocol
  • the SRIOV network card to be moved out of the virtual machine can be synchronized with the OVS network card in real time.
  • SRIOV NIC resources can be deactivated through cloud platform scripts.
  • the method may further include: synchronizing the service data of the OVS network card resource to be moved into the virtual machine to the SRIOV network card resource to be moved into the virtual machine, and The SRIOV NIC resources are activated.
  • the service routing data of the OVS NIC to be moved to the VM is synchronized to the SRIOV NIC to be moved to the VM so that the SRIOV NIC can take over the service.
  • the SRIOV network card can be activated through the cloud platform script to enable the SRIOV network card to take over the service.
  • the step of deploying the single input and output virtualized SIOOV network card resources in the network resource pool to the designated network virtual machine may include: following the PCI direct pass of the SIOOV NIC resources in the network resource pool. The mode is deployed to the network virtual machine.
  • the method may further include: switching the SRIOV network card resource of the virtual machine to be moved out from the PCI pass mode to the Macvtap pass mode.
  • the SRIOV NIC Since the SRIOV NIC does not support virtual machine hot migration in PCI Passthrough mode, the SRIOV NIC needs to be switched from PCI Passthrough mode to Macvtap Passthrough mode for hot migration of SRIOV NIC resources.
  • the method may further include: switching the SRIOV network card resource to be moved into the virtual machine from the Macvtap pass-through mode to the PCI pass-through mode.
  • the SRIOV NIC resources are migrated to the VM to be migrated in the Macvtap Passthrough mode, the SRIOV NIC resources are switched to the high-efficiency PCI Passthrough mode to ensure the virtual machine operating efficiency.
  • the step of collecting real-time service state data of all virtual machines in the network may include: collecting real-time transmission protocols of all virtual machines in the network according to a second preset period (Real- Time Transport Protocol (RTP) The number of packets, the number of lost packets, and the jitter duration.
  • RTP Real- Time Transport Protocol
  • the second preset period is smaller than the first preset period.
  • the service data of the virtual machine is periodically collected according to the second preset period, and the deployment of the SRIOV network card resources is periodically determined according to the first preset period.
  • the second preset period is smaller than the first preset period, so that at least one set of sampling data is collected in the decision period to determine the service running status of each virtual machine in the network, thereby determining whether to perform the migration of the SRIOV network card resources.
  • the second preset period as the sampling period can be set to be no less than 30 s.
  • the first preset period as the decision period may be set to be not less than 90 s.
  • the step of evaluating the current service running status of each virtual machine according to the real-time service status data collected in the first preset period may include: acquiring the current status of each virtual machine by using the following formula: Business Health Evaluation Parameter Si:
  • the Traf(i), Lost(i), and Jett(i) are respectively the number of RTP packets, the number of lost packets, and the jitter duration of the virtual machine i collected in the first preset period
  • Traf(avg) , Lost (avg), Jett (avg) are the average of the number of RTP packets of all the virtual machines collected in the first preset period, the average of the number of lost packets, and the average value of the jitter duration, respectively.
  • the number of RTP packets, the number of lost packets, and the jitter duration of the virtual machine are weighed to obtain the service running evaluation parameter Si of the virtual machine, which can effectively evaluate the service running status of each virtual machine.
  • Si the service running evaluation parameter
  • W1 and Wj may be set equal, and both are greater than Wt.
  • Wt, Wl, and Wj may be set equal to 1/3.
  • whether the device is in a busy condition may be determined according to the communication traffic within the preset time period of the device. For example, when the communication traffic of the device in the preset time is greater than the preset threshold, it is determined that the device is busy; when the communication traffic within the preset time is less than or equal to the preset threshold, it is determined that the device is not busy. In addition, it is also possible to determine whether the device is busy according to the degree of load during the operation of the device.
  • the device it is also possible to set the device to be in a busy state according to the membership data, and set the device to be busy in the time period; for example, setting the device to be busy at 6:30-9:00 and 17:00-19:00 The time period during which the device is considered to be busy regardless of the load status of the device, and the device that is considered to be in the remaining time period is not busy.
  • the step of acquiring the virtual machine to be moved in the virtual machine according to the current service running status of each virtual machine may include: evaluating the parameter Si according to the current service running status of each virtual machine, Obtaining a virtual machine to be moved into the virtual machine in the virtual machine, where the Si to be moved into the virtual machine is not less than the Si of the virtual machine to be moved out.
  • the value of Si to be moved into the virtual machine is not less than the value of Si to be moved out of the virtual machine.
  • the value of Si that can be moved into the virtual machine is not less than the value of Si other than the virtual machine to be moved into the virtual machine.
  • the number of selected virtual machines to be moved into is less than or equal to the number of virtual machines to be moved out.
  • the Si of each virtual machine in the network resource pool compares the Si of each virtual machine with the Si of the virtual machine to be moved out, when the Si of the virtual machine in the network resource pool is greater than or equal to the virtual machine to be moved out.
  • the method may include: obtaining, according to a current service running condition of each virtual machine, an amount equal to the number of the virtual machines to be moved out.
  • the virtual machine is to be moved into the virtual machine, and each of the virtual machines to be moved out is corresponding to one to be moved into the virtual machine, where the current service running status of the virtual machine to be moved into the virtual machine is not better than the to-be-moved The current business health of other virtual machines outside of the virtual machine.
  • the VMs to be moved out are selected to be moved to the virtual machine, and each VM to be moved out is corresponding to one to be moved.
  • the SRIOV NIC resources of each VM to be moved out are migrated to the corresponding VMs to be moved into the virtual machine, so that each SRIOV NIC resource is effectively adjusted in real time.
  • the current service running status of the virtual machine to be moved into the virtual machine is inferior to the current running status of the virtual machine except the virtual machine to be moved into the virtual machine, so that the SRIOV NIC resource is applied to the virtual machine with the worst running status. Improve the running status of the virtual machine to ensure load balancing.
  • the method for deploying network resources may be applied to a controller. As shown in FIG. 2, the method may include steps 201-207.
  • step 201 the VFA corresponding to the SRIOV network card PFA in the network resource pool is deployed to the virtual machine 1 (Virtual Machine 1, VM1) according to the PCI Passthrough mode.
  • the VFB corresponding to the SRIOV network card PFB is deployed to the VM2 in PCI Passthrough mode (referred to as PVFB), and the common network card resources of the OVS network card are separately deployed for each virtual machine VM1-VM5.
  • the virtual machines (VM1, VM2) on which the SRIOV network card and the OVS network card are deployed enable the SRIOV network card in the PCI Passthrough mode, the OVS network card is not enabled, and the SRIOV network card and the OVS network card are configured with the same IP routing information and are synchronized in real time.
  • step 202 the number of RTP packets, the number of lost packets, and the jitter duration of all virtual machines are collected according to the second preset period, and the second preset period is not less than 30 seconds.
  • step 203 according to the number of RTP packets, the number of lost packets, and the jitter duration of each virtual machine collected in the first preset period, the current service running evaluation parameter Si of each virtual machine is calculated according to the following formula:
  • the first preset period is not less than 90s.
  • Wl and Wj are equal, and both are greater than Wt; when the device is not busy, Wt, Wl, and Wj are equal, respectively equal to 1/3.
  • step 204 according to the current service running condition evaluation parameter Si of each virtual machine, two virtual machines to be moved into the virtual machine are obtained in all the virtual machines, wherein the Si to be moved into the virtual machine is not less than the Si of other virtual machines in the network. That is, the business to be moved into the virtual machine has the worst business performance.
  • the current service health evaluation parameter Si of the virtual machine The larger the value of the current service health evaluation parameter Si of the virtual machine, the worse the current service running status of the virtual machine is. Therefore, when two VMs to be moved into the virtual machine are acquired in all the virtual machines, the current service running state evaluation parameters Si of all the virtual machines can be compared. Assume that the Si of VM4 is the largest, and the Si of VM5 is the second. As shown in FIG. 3, VM4 and VM5 are regarded as being to be moved into the virtual machine. The virtual machines VM1 and VM2 to which the SRIOV network card resources are allocated are regarded as the virtual machines to be moved out, and VM1 corresponds to VM5, and VM2 corresponds to VM4.
  • step 205 as shown in FIG. 4, the SRIOV network card resources on VM1 and VM2 are deactivated by the cloud platform script, the OVS network card automatically takes over the service, and the SRIOV network card on VM1 and VM2 is switched to the Macvtap Passthrough mode, respectively. Recorded as MVFB and MVFA.
  • step 206 as shown in FIG. 5, the SRIOV network card resource MVFB of VM1 is migrated to VM5, and the SRIOV network card resource MVFA of VM2 is migrated to VM4.
  • step 207 the service data of the OVS network card of the VM4 is synchronized to the SRIOV network card MVFB, and the service data of the OVS network card of the VM5 is synchronized to the SRIOV network card MVFA.
  • the MVFA and MVFB are respectively switched to the PCI Passthrough mode, which is denoted as PVFA and PVFB. Activate PVFA and PVFB to automatically take over the SRIOV NIC on VM5 and VM4.
  • the method for deploying network resources in an embodiment firstly deploys an SRIOV network card in a virtual machine with a high historical load according to historical experience, and simultaneously deploys an OVS network card for each virtual machine, and then evaluates the current service running status of each virtual machine in the network, and finds that If the virtual machine with poor business performance uses the OVS common network card, the SRIOV network card is migrated to the virtual machine with poor business performance.
  • the method dynamically and reasonably deploys the SRIOV NIC resources according to the real-time service status of the virtual machine, so that the load of all the virtual machines in the network is balanced, and the utilization efficiency of the SRIOV NIC resources is improved. Therefore, the technical problem that the SRIOV network card cannot be deployed reasonably dynamically is solved.
  • the embodiment further provides an apparatus for deploying network resources, including an acquisition module, an acquisition module, and a migration module.
  • the collection module is configured to collect real-time service state data of all virtual machines of the network after the single input and output virtualized SIOOV network card resources in the network resource pool are deployed to the network virtual machine, and collect the data according to the first preset period.
  • the real-time business status data evaluates the current business health of each virtual machine.
  • the obtaining module is configured to obtain a virtual machine to be migrated into the virtual machine according to the current service running status of each virtual machine, and use the virtual machine that has deployed the SRIOV network card resource as the to-be-migrated virtual machine, where the The current service running status of the virtual machine is not better than the current service running status of the virtual machine to be moved out.
  • the migration module is configured to be different from the virtual machine to be moved out to the virtual machine to be moved out.
  • the SRIOV network card resource to be moved out of the virtual machine is migrated to the virtual machine to be moved into the virtual machine.
  • the network resource deployment device of the present embodiment dynamically and reasonably deploys the SRIOV network card resources according to the real-time service status of the virtual machine, so that the load of all the virtual machines in the network is balanced, and the utilization efficiency of the SRIOV network card resources is improved. It can solve the technical problem that the SRIOV network card cannot be deployed reasonably dynamically.
  • the collecting module may include: a first collecting unit.
  • the first collection unit is configured to deploy the SRIOV network card resources in the network resource pool to the pre-stated virtual machines with high load.
  • the device may further include a deployment module, a deactivation module, and an activation module.
  • the deployment module is configured to deploy open source virtual switching OVS NIC resources for each virtual machine of the network.
  • the deactivation module is configured to deactivate the SRIOV network card resources of the virtual machine to be moved out after the SIOVO network card resources of the virtual machine to be migrated are synchronized with the service data of the OVS network card resources of the virtual machine to be moved out. .
  • the activation module is configured to synchronize the service data of the OVS network card resource to be moved into the virtual machine to the SRIOV network card resource to be moved into the virtual machine, and activate the SRIOV network card resource to be moved into the virtual machine.
  • the collection module may include a deployment unit.
  • the deployment unit is configured to deploy the SRIOV NIC resources in the network resource pool to the network virtual machine in PCI passthrough mode.
  • the apparatus can also include a first switching module.
  • the first switching module is configured to switch the SRIOV network card resource of the virtual machine to be moved out from the PCI pass-through mode to the Macvtap pass-through mode.
  • the device may further include a second switching module.
  • the second switching module is configured to switch the SRIOV network card resource to be moved into the virtual machine from the Macvtap pass-through mode to the PCI pass-through mode.
  • the collection module may include a second collection unit.
  • the second collection unit is configured to collect real-time transmissions of all virtual machines of the network according to a second preset period.
  • the collecting module may further include a first acquiring unit.
  • the first obtaining unit is configured to obtain a current service running evaluation parameter Si of each virtual machine by using the following formula:
  • the Traf(i), Lost(i), and Jett(i) are respectively the number of RTP packets, the number of lost packets, and the jitter duration of the virtual machine i collected in the first preset period
  • Traf(avg) , Lost (avg), Jett (avg) are the average of the number of RTP packets of all the virtual machines collected in the first preset period, the average of the number of lost packets, and the average value of the jitter duration, respectively.
  • the obtaining module may include a second acquiring unit.
  • a second obtaining unit configured to obtain a parameter to be migrated into the virtual machine according to the current service running condition evaluation parameter Si of each virtual machine, where the Si to be moved into the virtual machine is not less than the to-be-sold Move out of the virtual machine Si.
  • the obtaining module may include a third acquiring unit.
  • the third obtaining unit is configured to acquire, according to the current service running status of each virtual machine, the virtual machines to be moved into the virtual machine that are equal to the number of the virtual machines to be moved out, and make each of the to-be-moved
  • the virtual machines respectively correspond to one of the virtual machines to be moved into the virtual machine, and the current service running status of the virtual machine to be moved into the virtual machine is not better than the current running status of the virtual machines except the virtual machine to be moved into the virtual machine.
  • the network resource deployment device of the embodiment dynamically and reasonably deploys the SRIOV network card resource according to the real-time service status of the virtual machine, so that the load of each virtual machine in the network is balanced, and the utilization efficiency of the SRIOV network card resource is improved. Therefore, the technical problem that the SRIOV network card resources cannot be reasonably and dynamically deployed can be solved.
  • the device for deploying the network resource is a device corresponding to the method for deploying the network resource, and all implementation manners in the foregoing method embodiments are applicable to the embodiment of the device.
  • the embodiment further provides a controller, including the device for deploying network resources as described in any of the above embodiments.
  • This embodiment also provides a core network device, including the controller as described in the above embodiments.
  • the embodiment further provides a computer readable storage medium storing computer executable instructions for performing any of the above methods for network resource deployment.
  • FIG. 9 is a schematic diagram of a general hardware structure of the core network device provided in this embodiment.
  • the general structure of the core network device may include: a processor 310 and a memory 320; A communication interface 330 and a bus 340 may also be included.
  • the processor 310, the memory 320, and the communication interface 330 can complete communication with each other through the bus 340. Communication interface 330 can be used for information transmission.
  • the processor 310 can invoke logic instructions in the memory 320 to perform the method of network resource deployment of the above embodiments.
  • the logic instructions in the memory 320 described above may be implemented in the form of a software functional unit and sold or used as a stand-alone product, and may be stored in a computer readable storage medium.
  • the technical solution of the present disclosure may be embodied in the form of a software product stored in a storage medium, including a plurality of instructions for causing a computer device (which may be a personal computer, a server, or a network) The device or the like) performs all or part of the steps of the method described in this embodiment.
  • the foregoing storage medium may be a non-transitory storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.
  • a medium that can store program code, or a transitory storage medium including: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.
  • the program when executed, may include a flow of an embodiment of the method described above, wherein the computer readable storage medium may be a magnetic disk, an optical disk, a read only memory (ROM), or a random access memory. (RAM), etc.
  • the computer readable storage medium may be a magnetic disk, an optical disk, a read only memory (ROM), or a random access memory. (RAM), etc.
  • the size of the sequence number of the process does not mean the order of execution, and the order of execution of each process should be determined by the function and internal logic of the process, and should not be implemented in this embodiment.
  • the process constitutes any qualification.
  • the present disclosure provides a method, an apparatus, a controller, and a core network device for deploying network resources, which can dynamically and reasonably deploy SRIOV NIC resources in a virtual machine, so that load balancing of all virtual machines in the network is balanced, and utilization of SRIOV NIC resources is improved. effectiveness. Therefore, the technical problem that the SRIOV network card cannot be reasonably and dynamically deployed can be solved.

Landscapes

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

Abstract

一种网络资源部署的方法、装置、控制器及核心网设备，该方法包括：将SRIOV网卡资源部署到指定的网络虚拟机后，采集网络中所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况；将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，根据每个虚拟机的当前业务运行状况确定待迁入虚拟机，其中待迁入虚拟机的当前业务运行状况劣于待迁出虚拟机的当前业务运行状况，且所述待迁入虚拟机的数量小于或等于待迁出虚拟机的数量；将待迁出虚拟机的SRIOV网卡资源迁移到待迁入虚拟机中。

Description

网络资源部署的方法、装置、控制器及核心网设备 技术领域
本公开涉及网络虚拟技术领域，例如涉及一种网络资源部署的方法、装置、控制器及核心网设备。
背景技术
开源虚拟交换(Open Virtualized Switch，OVS)是常见的网络资源的虚拟化技术，它可以遵从开源流Open Flow协议，通过软件手段，在虚拟化管理层将硬件网络资源虚拟化以便应用层统一使用。其中网络资源的计算和部署的开销主要体现在虚拟化管理层。
单根输入输出虚拟化(Single Root I/O Virtualization，SRIOV)可以用来解决虚拟共享的问题，即多个虚拟机可以同时共享同一个外围部件接口(Peripheral Component Interconnect，PCI)硬件，可以让应用层的虚拟功能(Virtualized Function，VF)直接使用PCI硬件中对应的物理功能(Physical Function，PF)，以提高效率。其中网络资源的计算和部署开销主要在硬件设备上，它需要与门支持SRIOV的硬件网卡。
SRIOV可以包括PCI直通(Passthrough)模式和Macvtap直通(Passthrough)模式。Macvtap Passthrough模式需要经过中间模拟层转发，效率没有PCI Passthrough模式高，但可以支持虚拟机热迁移，而PCI Passthrough模式不支持虚拟机热迁移。
在应用性能上看，SRIOV网卡虽然种种性能都比OVS网卡好，但是SRIOV网卡需要特定的硬件资源，造价高，对于网络状况好的环境性价比很低。所以对于一个虚拟化环境来说，SRIOV网卡要部署在环境中网络负荷高的虚拟机上，但是目前的网络规划都是将SRIOV网卡静态部署到特定的虚拟机上，无法合理动态地部署SRIOV网卡。
发明内容
本公开提供一种网络资源部署的方法、装置、控制器及核心网设备，可以解决无法合理动态地部署SRIOV网卡的技术问题。
本实施例提供一种网络资源部署的方法，可以包括：在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机后，采集所述网络中所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况。将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，根据每个虚拟机的当前业务运行状况确定待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况劣于所述待迁出虚拟机的当前业务运行状况，且所述待迁入虚拟机的数量小于或等于待迁出虚拟机的数量。将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中。
可选地，所述将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机的步骤可以包括：将网络资源池中的SRIOV网卡资源部署到预先统计出的负荷高于负荷阈值的虚拟机。
可选地，所述在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机后，所述方法还可以包括：为所述网络中的所有虚拟机分别部署开源虚拟交换OVS网卡资源。
所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，所述方法还可以包括：在所述待迁出虚拟机的SRIOV网卡资源与所述待迁出虚拟机的OVS网卡资源的业务数据同步之后，对所述待迁出虚拟机的SRIOV网卡资源进行去激活。
所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之后，所述方法还可以包括：将所述待迁入虚拟机的OVS网卡资源的业务数据同步给所述待迁入虚拟机的SRIOV网卡资源，并对所述待迁入虚拟机的SRIOV网卡资源进行激活。
可选地，所述将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机的步骤可以包括：将网络资源池中的SRIOV网卡资源按照PCI直通模式部署到指定的网络虚拟机。
所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，所述方法还可以包括：将所述待迁出虚拟机的SRIOV网卡资源从PCI直通模式切换到Macvtap直通模式。
可选地，所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之后，所述方法还可以包括：将所述待迁入虚拟机的SRIOV网卡资源从Macvtap直通模式切换到PCI直通模式。
可选地，所述采集所述网络中所有虚拟机的实时业务状态数据的步骤包括：
按照第二预设周期采集所述网络中所有虚拟机的实时传输协议RTP报文数、丢包数和抖动时长，所述第二预设周期小于所述第一预设周期。
可选地，所述根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况的步骤可以包括：通过如下公式获取每个虚拟机的当前业务运行状况评价参数Si：
Figure PCTCN2017080907-appb-000001
可选地，Traf(i)、Lost(i)、Jett(i)分别为在所述第一预设周期内采集到的虚拟机i的RTP报文数、丢包数和抖动时长，Traf(avg)、Lost(avg)、Jett(avg)分别为在所述第一预设周期内采集到的所有虚拟机的RTP报文数的平均值、丢包数的平均值和抖动时长的平均值，Wt、Wl、Wj分别为预设的RTP报文数评价权重、丢包数评价权重和抖动时长评价权重，Wt+Wl+Wj＝1。
所述根据每个虚拟机的当前业务运行状况，确定待迁入虚拟机的步骤可以包括：根据每个虚拟机的当前业务运行状况评价参数Si，确定待迁入虚拟机，其中所述待迁入虚拟机的Si大于或等于所述待迁出虚拟机的Si。
可选地，所述根据网络中每个虚拟机的当前业务运行状况，确定待迁入虚拟机的步骤可以包括：根据网络中每个虚拟机的当前业务运行状况，确定在所 述所有虚拟机中获取与所述待迁出虚拟机数量相等的待迁入虚拟机，并使每个所述待迁出虚拟机对应一个所述待迁入虚拟机。
本实施例还提供了一种网络资源部署的装置，可以包括：采集模块、获取模块和迁移模块。
采集模块，设置为在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机后，采集所述网络中所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况。
获取模块，设置为将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，根据每个虚拟机的当前业务运行状况，确定待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况劣于所述待迁出虚拟机的当前业务运行状况，且所述待迁入虚拟机的数量小于或等于待迁出虚拟机的数量。
迁移模块，设置为将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中。
可选地，所述装置还可以包括：部署模块，设置为为所述网络的所有虚拟机分别部署开源虚拟交换OVS网卡资源。
去激活模块，设置为在采集模块将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，在所述待迁出虚拟机的SRIOV网卡资源与所述待迁出虚拟机的OVS网卡资源的业务数据同步之后，对所述待迁出虚拟机的SRIOV网卡资源进行去激活。
激活模块，设置为在迁移模块将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之后，将所述待迁入虚拟机的OVS网卡资源的业务数据同步给所述待迁入虚拟机的SRIOV网卡资源，并对所述待迁入虚拟机的SRIOV网卡资源进行激活。
可选地，所述采集模块可以包括：部署单元，设置为将网络资源池中的SRIOV网卡资源按照PCI直通模式部署到网络虚拟机。
可选地，所述装置还可以包括：第一切换模块，设置为在迁移模块将所述 待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，将所述待迁出虚拟机的SRIOV网卡资源从PCI直通模式切换到Macvtap直通模式。
本实施例还提供一种控制器，包括：如上任一项所述的网络资源部署的装置。
本实施例还提供一种核心网设备，包括：如上的控制器。
本实施例还提供一种计算机可读存储介质，存储有计算机可执行指令，所述计算机可执行指令用于执行上述任意一种网络资源部署的方法。
本实施例还提供一种核心网设备，该核心网设备包括一个或多个处理器、存储器以及一个或多个程序，所述一个或多个程序存储在存储器中，当被一个或多个处理器执行时，执行上述任意一种网络资源部署的方法。
本实施例还提供了一种计算机程序产品，所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算机程序，所述计算机程序包括程序指令，当所述程序指令被计算机执行时，使所述计算机执行上述任意一种网络资源部署的方法。
本公开的网络资源部署的方法，根据网络中每个虚拟机的实时业务状况，动态合理地在虚拟机中部署SRIOV网卡资源，可以达到网络中所有虚拟机的负荷均衡，提高SRIOV网卡资源的利用效率。从而可以解决无法合理动态地部署SRIOV网卡的技术问题。
附图说明
图1为本实施例网络资源部署的方法流程图。
图2为本实施例网络资源部署的方法的应用实例流程图。
图3为本实施例网络资源部署的方法网卡资源初始部署对应的网络结构示意图。
图4为本实施例网络资源部署的方法网卡资源模式切换对应的网络结构示意图。
图5为本实施例网络资源部署的方法网卡资源迁移对应的网络结构示意图。
图6为本实施例网络资源部署的方法网卡资源数据同步对应的网络结构示意图。
图7为本实施例网络资源部署的方法网卡资源切换启用对应的网络结构示意图。
图8为本实施例网络资源部署的装置的结构示意图。
图9为本实施例的核心网设备的通用硬件结构示意图。
具体实施方式
为使本公开要解决的技术问题、技术方案更加清楚，下面将结合附图及实施例进行描述。在不冲突的情况下，以下实施例和实施例中的技术方案可以相互结合。
如图1所示，本实施例的网络资源部署的方法可以包括：步骤110-步骤130。
在步骤110中，在将网络资源池中的SRIOV网卡资源部署到指定的网络虚拟机后，采集所述网络中所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的实时业务状态数据评价每个虚拟机的当前业务运行状况。
通过采集网络中每个虚拟机的实时业务状态数据，可有效评价每个虚拟机的当前业务运行状况，从而根据虚拟机的业务运行状况合理、动态地调整SRIOV网卡资源。
虚拟机的当前业务运行状况可以代表虚拟机当前的负荷情况，一般虚拟机当前的负荷越高则当前业务运行状况越差。
此外，虚拟机的当前业务运行状况也可以代表虚拟机当前所处的网络质量，或虚拟机发送和接收数据的时延或丢包率等参数。
在步骤120中，将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，根据每个虚拟机的当前业务运行状况，在所述所有虚拟机中确定待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况劣于或等于所述待迁出虚拟机的当前业务运行状况。
基于虚拟机的当前业务运行状况选取待迁入虚拟机，从而基于虚拟机的业 务运行状况动态调整SRIOV网卡资源。
可选地，可将网络中所有虚拟机的当前业务运行状态与已经部署了SRIOV网卡资源的待迁出虚拟机的当前业务运行状态进行比较，确定比已经部署了SRIOV网卡资源的待迁出虚拟机的当前业务运行状态差或相等的虚拟机，并将该虚拟机作为待迁入虚拟机，即该待迁入虚拟机可以是未部署SRIOV网卡资源的虚拟机，也可以是部署了SRIOV网卡资源的虚拟机。
可选地，在网络的所有虚拟机中，比较每个虚拟机的当前业务运行状态，将当前业务运行状态最差的虚拟机最为待迁入虚拟机，即待迁入虚拟机的当前业务运行状况劣于网络中除待迁入虚拟机之外的其他虚拟机的当前业务运行状况。
在步骤130中，在所述待迁入虚拟机与所述待迁出虚拟机为不同的虚拟机时，将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中。
通过将待迁出虚拟机中的SRIOV网卡资源迁移到业务运行状况差的待迁入虚拟机中，能够改善待迁入虚拟机的业务运行状况，从而可以保证网络中每个虚拟机可稳定运行，使得网络中每个虚拟机的负荷均衡。
本实施例的网络资源部署的方法，根据虚拟机的实时业务状况，动态合理地部署SRIOV网卡资源，可以达到网络中所有虚拟机的负荷均衡，提高SRIOV网卡资源的利用效率。可以解决无法合理动态地部署SRIOV网卡的技术问题。
可选地，在步骤110中，所述将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到网络虚拟机的步骤可以包括：将网络资源池中的SRIOV网卡资源部署到预先统计出的负荷高的虚拟机。
可选地，预先计算网络资源池中所有虚拟机的负荷，选出负荷高于负荷阈值的虚拟机，将并SRIOV网卡资源部署到选出的虚拟机中。其中，该负荷阈值可以是工作人员根据经验值确定的，也可以是系统根据网络资源池中所有虚拟机的历史负荷数据计算得到的。
可按照历史经验在历史负荷高于负荷阈值的虚拟机部署SRIOV网卡资源，从而优化初始部署，减少后续SRIOV网卡资源动态调整的次数，以节省动态调 整的资源。
可选地，在步骤110中，所述在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到网络虚拟机后，所述方法还可以包括：为所述网络中的所有虚拟机分别部署OVS网卡资源。
为网络中的所有虚拟机分别部署OVS网卡资源。对于同时具有SRIOV网卡资源和OVS网卡资源的虚拟机，由于SRIOV网卡优先于OVS网卡启用，故虚拟机上启用的是SRIOV网卡，另外OVS网卡的容量不受硬件限制，所以配置一个未被启用的OVS网卡不影响虚拟机的性能和业务。对于仅部署OVS网卡资源的虚拟机，则启用OVS普通网卡进行工作。
在步骤130之前，所述方法还可以包括：在所述待迁出虚拟机的SRIOV网卡资源与所述待迁出虚拟机的OVS网卡资源的业务数据同步之后，对所述待迁出虚拟机的SRIOV网卡资源进行去激活。
可选地，若检测到网络中未部署SRIOV网卡资源的虚拟机的业务运行状况比已经部署了SRIOV网卡资源的虚拟机的业务运行状况差，则需要进行SRIOV网卡资源的迁移。在迁移之前，需要保证待迁出虚拟机上的SRIOV网卡与OVS网卡的业务数据同步，即配置的网络互连协议(Internet Protocol，IP)路由信息完全相同，然后对待迁出虚拟机的SRIOV网卡资源进行去激活，使待迁出虚拟机上的OVS网卡接管业务。
可选地，可使待迁出虚拟机的SRIOV网卡与OVS网卡进行实时同步。可通过云平台脚本对SRIOV网卡资源进行去激活。
在步骤130之后，所述方法还可以包括：将所述待迁入虚拟机的OVS网卡资源的业务数据同步给所述待迁入虚拟机的SRIOV网卡资源，并对所述待迁入虚拟机的SRIOV网卡资源进行激活。
将SRIOV网卡资源迁移到待迁入虚拟机之后，需要将待迁入虚拟机的OVS网卡的业务路由数据同步给该待迁入虚拟机的SRIOV网卡，以便SRIOV网卡能够接管业务。可通过云平台脚本对SRIOV网卡进行激活，使SRIOV网卡接管业务。
通过为网络中的所有虚拟机分别部署OVS网卡资源，可以保证未部署SRIOV网卡的虚拟机以及SRIOV网卡迁移后的虚拟机的业务的正常运行。
可选地，在步骤110中，所述将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机的步骤可以包括：将网络资源池中的SRIOV网卡资源按照PCI直通模式部署到网络虚拟机。
按照PCI直通PCI Passthrough模式部署SRIOV网卡资源，可以保证网络虚拟机的运行效率。
在步骤130之前，所述方法还可以包括：将所述待迁出虚拟机的SRIOV网卡资源从PCI直通模式切换到Macvtap直通模式。
由于SRIOV网卡在PCI Passthrough模式不支持虚拟机热迁移，因此需要将SRIOV网卡从PCI Passthrough模式切换到Macvtap直通(Macvtap Passthrough)模式，以对SRIOV网卡资源进行热迁移。
可选地，在步骤130之后，所述方法还可以包括：将所述待迁入虚拟机的SRIOV网卡资源从Macvtap直通模式切换到PCI直通模式。
将SRIOV网卡资源以Macvtap Passthrough模式迁移到待迁入虚拟机之后，将SRIOV网卡资源切换到高效率的PCI Passthrough模式，可以保证虚拟机的运行效率。
可选地，在步骤110中，所述采集所述网络中所有虚拟机的实时业务状态数据的步骤可以包括：按照第二预设周期采集所述网络中所有虚拟机的实时传输协议(Real-time Transport Protocol，RTP)报文数、丢包数和抖动时长，所述第二预设周期小于所述第一预设周期。
按照第二预设周期定期采集虚拟机的业务数据，按照第一预设周期定期决策SRIOV网卡资源的部署。第二预设周期小于第一预设周期，从而可以保证决策周期内至少采集一组采样数据，以判断网络中每个虚拟机的业务运行状况，从而决策是否进行SRIOV网卡资源的迁移。
为了保证采样数据的实时性，可设定作为采样周期的第二预设周期不小于30s。为了保证不乒乓迁移，可设定作为决策周期的第一预设周期不小于90s。
可选地，在步骤110中，所述根据第一预设周期内采集到的实时业务状态数据评价每个虚拟机的当前业务运行状况的步骤可以包括：通过如下公式获取每个虚拟机的当前业务运行状况评价参数Si：
Figure PCTCN2017080907-appb-000002
其中，Traf(i)、Lost(i)、Jett(i)分别为在所述第一预设周期内采集到的虚拟机i的RTP报文数、丢包数和抖动时长，Traf(avg)、Lost(avg)、Jett(avg)分别为在所述第一预设周期内采集到的所有虚拟机的RTP报文数的平均值、丢包数的平均值和抖动时长的平均值，Wt、Wl、Wj分别为预设的RTP报文数评价权重、丢包数评价权重和抖动时长评价权重，Wt+Wl+Wj＝1。
通过上述公式权衡虚拟机的RTP报文数、丢包数和抖动时长，得到虚拟机的业务运行状况评价参数Si，可有效评估每个虚拟机的业务运行状况。其中Si的值越大，则表明虚拟机上的业务运行状况越差。
可选地，设备运行忙时，考虑到保证网络的传输质量，可设定Wl和Wj相等，且均大于Wt。设备运行非忙时，可设定Wt、Wl和Wj相等，分别等于1/3。
可选地，可根据设备预设时间段内的通信流量确定设备是否处于繁忙状况。例如，当设备在预设时间内的通信流量大于预设阈值时，确定设备运行忙；当在预设时间内的通信流量小于或等于预设阈值时，确定设备运行非忙。此外，还可以根据对设备运行过程中的负荷程度确定设备是否处于繁忙状态。也可以根据隶属数据，将设备使用频繁的时间段，将该时间段内的设备设定为处于繁忙状态；例如，设定6:30-9:00以及17:00-19:00为设备繁忙时间段，在该时间段内，无论设备的负荷状况如何，均认为设备运行忙，而认为处于其余时间段的设备运行非忙。
在步骤120中，所述根据每个虚拟机的当前业务运行状况，在所述所有虚拟机中获取待迁入虚拟机的步骤可以包括：根据每个虚拟机的当前业务运行状况评价参数Si，在所述所有虚拟机中获取待迁入虚拟机，其中所述待迁入虚拟机的Si不小于所述待迁出虚拟机的Si。
由于Si的值越大，虚拟机上的业务运行状况越差，所以选择待迁入虚拟机的Si的值不小于待迁出虚拟机的Si的值。可选择待迁入虚拟机的Si的值不小于除待迁入虚拟机之外的其他虚拟机的Si的值。
可选地，选出的待迁入虚拟机的数量小于或等于待迁出虚拟机的数量。在计算得到网络资源池中每个虚拟机的Si后，将每个虚拟机的Si与待迁出虚拟机的Si进行比较，当网络资源池中虚拟机的Si大于或等于待迁出虚拟机的Si的30％时，确定该虚拟机为待迁入虚拟机。例如，待迁出虚拟机的Si＝35，网络资源池中一个虚拟机的Si＝70，此时，(70-35)/70＝50％，该虚拟机的Si大于待迁出虚拟机的Si的30％，则确定该虚拟机为待迁入虚拟机，将待迁出虚拟机中的SRIOV资源迁入到该待迁入虚拟机中。
可选地，对于多个SRIOV网卡资源的情况，在步骤120中可以包括：根据每个虚拟机的当前业务运行状况，在所述所有虚拟机中获取与所述待迁出虚拟机数量相等的待迁入虚拟机，并使每个所述待迁出虚拟机分别对应一个所述待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况不优于除所述待迁入虚拟机之外的其他虚拟机的当前业务运行状况。
可选地，将多个SRIOV网卡资源分配给多个待迁出虚拟机之后，选择与待迁出虚拟机数量相等的待迁入虚拟机，并使每个待迁出虚拟机对应一个待迁入虚拟机，从而将每个待迁出虚拟机的SRIOV网卡资源迁移到与之对应的待迁入虚拟机中，保证每个SRIOV网卡资源都得到有效的实时动态调整。
其中，待迁入虚拟机的当前业务运行状况劣于除待迁入虚拟机之外的其他虚拟机的当前业务运行状况，使SRIOV网卡资源运用到业务运行状况最差的虚拟机中，以有效提升虚拟机的运行状况，保证负荷均衡。
下面对本实施例的应用实例，在该应用实例中，网络资源部署的方法可以应用于一控制器(Controller)，如图2所示，该方法可以包括步骤201-步骤207。
在步骤201中，如图3所示，将网络资源池中的SRIOV网卡PFA对应的VFA按照PCI Passthrough模式部署到虚拟机1(Virtual Machine 1，VM1)上(记 为PVFA)。将SRIOV网卡PFB对应的VFB按照PCI Passthrough模式部署到VM2上(记为PVFB)，并为每个虚拟机VM1-VM5分别部署OVS网卡的普通网卡资源。
此外，部署了SRIOV网卡和OVS网卡的虚拟机(VM1、VM2)按照PCI Passthrough模式启用SRIOV网卡，OVS网卡不启用，且SRIOV网卡与OVS网卡配置相同的IP路由信息，并实时同步。
在步骤202中，按照第二预设周期采集所有虚拟机的RTP报文数、丢包数和抖动时长，第二预设周期不小于30s。
在步骤203中，根据第一预设周期内采集的每个虚拟机的RTP报文数、丢包数和抖动时长，按照如下公式计算每个虚拟机的当前业务运行状况评价参数Si：
Figure PCTCN2017080907-appb-000003
其中，Traf(i)、Lost(i)、Jett(i)分别为在第一预设周期内采集到的虚拟机i的RTP报文数、丢包数和抖动时长，Traf(avg)、Lost(avg)、Jett(avg)分别为在第一预设周期内采集到的虚拟机的RTP报文数的平均值、丢包数的平均值和抖动时长的平均值，Wt、Wl、Wj分别为预设的RTP报文数评价权重、丢包数评价权重和抖动时长评价权重，Wt+Wl+Wj＝1。
可选地，第一预设周期不小于90s。设备运行忙时，Wl和Wj相等，且均大于Wt；设备运行非忙时，Wt、Wl和Wj相等，分别等于1/3。
在步骤204中，根据每个虚拟机的当前业务运行状况评价参数Si，在所有虚拟机中获取两个待迁入虚拟机，其中待迁入虚拟机的Si不小于网络中其他虚拟机的Si，即待迁入虚拟机的业务运行状况最差。
虚拟机的当前业务运行状况评价参数Si的值越大，表明该虚拟机的当前业务运行状况越差。因此，在所有虚拟机中获取两个待迁入虚拟机时，可将所有虚拟机的当前业务运行状态评价参数Si进行比较。假设VM4的Si最大，VM5的Si次之，如图3所示，将VM4和VM5作为待迁入虚拟机。将分配有SRIOV网卡资源的虚拟机VM1和VM2作为待迁出虚拟机，且VM1与VM5对应，VM2与VM4对应。
在步骤205中，如图4所示，将VM1和VM2上的SRIOV网卡资源通过云平台脚本去激活，OVS网卡会自动接管业务，并将VM1和VM2上的SRIOV网卡切换成Macvtap Passthrough模式，分别记为MVFB和MVFA。
在步骤206中，如图5所示，将VM1的SRIOV网卡资源MVFB迁移到VM5上，将VM2的SRIOV网卡资源MVFA迁移到VM4上。
在步骤207中，如图6所示，将VM4的OVS网卡的业务数据同步给SRIOV网卡MVFB，将VM5的OVS网卡的业务数据同步给SRIOV网卡MVFA。如图7所示，将MVFA和MVFB分别切换成PCI Passthrough模式，记为PVFA和PVFB。激活PVFA和PVFB，使VM5和VM4上的SRIOV网卡自动接管业务。
实施例的网络资源部署的方法，首先按照历史经验在历史负荷高的虚拟机部署SRIOV网卡，同时为每个虚拟机部署OVS网卡，然后评价网络中每个虚拟机的当前业务运行状况，发现有业务运行状况差的虚拟机使用的是OVS普通网卡，则将SRIOV网卡迁移到业务运行状况差的虚拟机上。该方法根据虚拟机的实时业务状况，动态合理地部署SRIOV网卡资源，可以使得网络中所有虚拟机的负荷均衡，提高SRIOV网卡资源的利用效率。从而解决无法合理动态地部署SRIOV网卡的技术问题。
如图8所示，本实施例还提供一种网络资源部署的装置，包括采集模块、获取模块和迁移模块。
采集模块设置为在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到网络虚拟机后，采集所述网络所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的实时业务状态数据评价每个虚拟机的当前业务运行状况。
获取模块设置为根据每个虚拟机的当前业务运行状况，在所述所有虚拟机中获取待迁入虚拟机，并将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，其中所述待迁入虚拟机的当前业务运行状况不优于所述待迁出虚拟机的当前业务运行状况。
迁移模块设置为在所述待迁入虚拟机与所述待迁出虚拟机为不同的虚拟机 时，将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中。
本实施例的网络资源部署的装置，根据虚拟机的实时业务状况，动态合理地部署SRIOV网卡资源，可以使得网络中所有虚拟机的负荷均衡，提高SRIOV网卡资源的利用效率。可以解决无法合理动态地部署SRIOV网卡的技术问题。
可选地，所述采集模块可以包括：第一采集单元。
第一采集单元设置为将网络资源池中的SRIOV网卡资源部署到预先统计出的负荷高的虚拟机。
可选地，所述装置还可以包括部署模块、去激活模块和激活模块。
部署模块设置为为所述网络的每个虚拟机分别部署开源虚拟交换OVS网卡资源。
去激活模块设置为在所述待迁出虚拟机的SRIOV网卡资源与所述待迁出虚拟机的OVS网卡资源的业务数据同步之后，对所述待迁出虚拟机的SRIOV网卡资源进行去激活。
激活模块设置为将所述待迁入虚拟机的OVS网卡资源的业务数据同步给所述待迁入虚拟机的SRIOV网卡资源，并对所述待迁入虚拟机的SRIOV网卡资源进行激活。
可选地，所述采集模块可以包括部署单元。
部署单元设置为将网络资源池中的SRIOV网卡资源按照PCI直通模式部署到网络虚拟机。
所述装置还可以包括第一切换模块。
第一切换模块设置为将所述待迁出虚拟机的SRIOV网卡资源从PCI直通模式切换到Macvtap直通模式。
可选地，所述装置还可以包括第二切换模块。
第二切换模块设置为将所述待迁入虚拟机的SRIOV网卡资源从Macvtap直通模式切换到PCI直通模式。
可选地，所述采集模块可以包括第二采集单元。
第二采集单元设置为按照第二预设周期采集所述网络所有虚拟机的实时传 输协议RTP报文数、丢包数和抖动时长，所述第二预设周期小于所述第一预设周期。
可选地，所述采集模块还可以包括第一获取单元。
第一获取单元设置为通过如下公式获取每个虚拟机的当前业务运行状况评价参数Si：
Figure PCTCN2017080907-appb-000004
其中，Traf(i)、Lost(i)、Jett(i)分别为在所述第一预设周期内采集到的虚拟机i的RTP报文数、丢包数和抖动时长，Traf(avg)、Lost(avg)、Jett(avg)分别为在所述第一预设周期内采集到的所有虚拟机的RTP报文数的平均值、丢包数的平均值和抖动时长的平均值，Wt、Wl、Wj分别为预设的RTP报文数评价权重、丢包数评价权重和抖动时长评价权重，Wt+Wl+Wj＝1。
所述获取模块可以包括第二获取单元。
第二获取单元，设置为根据每个虚拟机的当前业务运行状况评价参数Si，在所述所有虚拟机中获取待迁入虚拟机，其中所述待迁入虚拟机的Si不小于所述待迁出虚拟机的Si。
可选地，所述获取模块可以包括第三获取单元。
第三获取单元设置为根据每个虚拟机的当前业务运行状况，在所述所有虚拟机中获取与所述待迁出虚拟机数量相等的待迁入虚拟机，并使每个所述待迁出虚拟机分别对应一个所述待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况不优于除所述待迁入虚拟机之外的其他虚拟机的当前业务运行状况。
本实施例的网络资源部署的装置，根据虚拟机的实时业务状况，动态合理地部署SRIOV网卡资源，可以使得网络中每个虚拟机的负荷均衡，提高SRIOV网卡资源的利用效率。从而可以解决无法合理动态地部署SRIOV网卡资源的技术问题。
需要说明的是，该网络资源部署的装置是与上述网络资源部署的方法相对应的装置，上述方法实施例中所有实现方式均适用于该装置的实施例中。
本实施例还提供了一种控制器，包括如上述实施例中任意一种所述的网络资源部署的装置。
本实施例还提供了一种核心网设备，包括如上述实施例中所述的控制器。
本实施例还提供一种计算机可读存储介质，存储有计算机可执行指令，所述计算机可执行指令用于执行上述任意一种网络资源部署的方法。
如图9所示，是本实施例提供的核心网设备的通用硬件结构示意图，如图9所示，该核心网设备的通用结构可以包括：处理器(processor)310和存储器(memory)320；还可以包括通信接口(Communications Interface)330和总线340。
其中，处理器310、存储器320和通信接口330可以通过总线340完成相互间的通信。通信接口330可以用于信息传输。处理器310可以调用存储器320中的逻辑指令，以执行上述实施例的网络资源部署的方法。
此外，上述的存储器320中的逻辑指令可以通过软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本公开的技术方案可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本实施例所述方法的全部或部分步骤。而前述的存储介质可以是非暂态存储介质，包括：U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质，也可以是暂态存储介质。
最后需要说明的是，本领域普通技术人员可理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来执行相关的硬件来完成的，该程序可存储于一个非暂态计算机可读存储介质中，该程序在执行时，可包括如上述方法的实施例的流程，其中，该计算机可读存储介质可以为磁碟、光盘、只读存储记忆体(ROM)或随机存储记忆体(RAM)等。
在上述实施例中，过程的序号的大小并不意味着执行顺序的先后，每个过程的执行顺序应以该过程的功能和内在逻辑确定，而不应对本实施例的实施过 程构成任何限定。
工业实用性
本公开提供了一种网络资源部署的方法、装置、控制器和核心网设备，可以动态合理地在虚拟机中部署SRIOV网卡资源，使得网络中所有虚拟机的负荷均衡，提高SRIOV网卡资源的利用效率。从而可以解决无法合理动态地部署SRIOV网卡的技术问题。

Claims (14)

  1. 一种网络资源部署的方法，包括：
    在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机后，采集所述网络中所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况；
    将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，根据网络中每个虚拟机的当前业务运行状况确定待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况劣于所述待迁出虚拟机的当前业务运行状况，且待迁入虚拟机的数量小于或等于待迁出虚拟机的数量；以及
    将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中。
  2. 根据权利要求1所述的方法，其中，所述将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机的步骤包括：
    将网络资源池中的SRIOV网卡资源部署到预先统计出的负荷高于负荷阈值的虚拟机中。
  3. 根据权利要求1所述的方法，所述在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机后，还包括：
    为所述网络中的所有虚拟机分别部署开源虚拟交换OVS网卡资源；
    所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，还包括：
    在所述待迁出虚拟机的SRIOV网卡资源与所述待迁出虚拟机的OVS网卡资源的业务数据同步之后，对所述待迁出虚拟机的SRIOV网卡资源进行去激活；以及
    所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之后，还包括：
    将所述待迁入虚拟机的OVS网卡资源的业务数据同步给所述待迁入虚拟机的SRIOV网卡资源，并对所述待迁入虚拟机的SRIOV网卡资源进行激活。
  4. 根据权利要求1所述的方法，其中，所述将网络资源池中的单根输入输 出虚拟化SRIOV网卡资源部署到指定的网络虚拟机的步骤包括：
    将网络资源池中的SRIOV网卡资源按照外围部件接口PCI直通模式部署到指定的网络虚拟机；
    所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，还包括：
    将所述待迁出虚拟机的SRIOV网卡资源从PCI直通模式切换到Macvtap直通模式。
  5. 根据权利要求4所述的方法，所述将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之后，还包括：
    将所述待迁入虚拟机的SRIOV网卡资源从Macvtap直通模式切换到PCI直通模式。
  6. 根据权利要求1所述的方法，其中，所述采集所述网络所有虚拟机的实时业务状态数据的步骤包括：
    按照第二预设周期采集所述网络中所有虚拟机的实时传输协议RTP报文数、丢包数和抖动时长，所述第二预设周期小于所述第一预设周期。
  7. 根据权利要求6所述的方法，其中，所述根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况的步骤包括：
    通过如下公式获取每个虚拟机的当前业务运行状况评价参数Si：
    Figure PCTCN2017080907-appb-100001
    其中，Traf(i)、Lost(i)、Jett(i)分别为在所述第一预设周期内采集到的虚拟机i的RTP报文数、丢包数和抖动时长，Traf(avg)、Lost(avg)、Jett(avg)分别为在所述第一预设周期内采集到的所有虚拟机的RTP报文数的平均值、丢包数的平均值和抖动时长的平均值，Wt、Wl、Wj分别为预设的RTP报文数评价权重、丢包数评价权重和抖动时长评价权重，Wt+Wl+Wj＝1；
    所述根据网络中每个虚拟机的当前业务运行状况确定待迁入虚拟机的步骤 包括：
    根据网络中每个虚拟机的当前业务运行状况评价参数确定待迁入虚拟机，其中所述待迁入虚拟机的Si大于或等于所述待迁出虚拟机的Si。
  8. 根据权利要求1所述的方法，其中，所述根据网络中每个虚拟机的当前业务运行状况，确定待迁入虚拟机的步骤包括：
    根据网络中每个虚拟机的当前业务运行状况，确定与所述待迁出虚拟机数量相等的待迁入虚拟机，并使每个所述待迁出虚拟机对应一个所述待迁入虚拟机。
  9. 一种网络资源部署的装置，包括：
    采集模块，设置为在将网络资源池中的单根输入输出虚拟化SRIOV网卡资源部署到指定的网络虚拟机后，采集所述网络中所有虚拟机的实时业务状态数据，并根据第一预设周期内采集到的每个虚拟机的实时业务状态数据评价每个虚拟机的当前业务运行状况；
    获取模块，设置为将已部署SRIOV网卡资源的虚拟机作为待迁出虚拟机，根据网络中每个虚拟机的当前业务运行状况确定待迁入虚拟机，其中所述待迁入虚拟机的当前业务运行状况劣于所述待迁出虚拟机的当前业务运行状况，且所述待迁入虚拟机的数量小于或等于待迁出虚拟机的数量；以及
    迁移模块，设置为将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中。
  10. 根据权利要求9所述的装置，还包括：
    部署模块，设置为为所述网络中的所有虚拟机分别部署开源虚拟交换OVS网卡资源；
    去激活模块，设置为在采集模块将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，在所述待迁出虚拟机的SRIOV网卡资源与所述待迁出虚拟机的OVS网卡资源的业务数据同步之后，对所述待迁出虚拟机的SRIOV网卡资源进行去激活；以及
    激活模块，设置为在迁移模块将所述待迁出虚拟机的SRIOV网卡资源迁移 到所述待迁入虚拟机中之后，将所述待迁入虚拟机的OVS网卡资源的业务数据同步给所述待迁入虚拟机的SRIOV网卡资源，并对所述待迁入虚拟机的SRIOV网卡资源进行激活。
  11. 根据权利要求9所述的装置，其中，所述采集模块包括：
    部署单元，设置为将网络资源池中的SRIOV网卡资源按照外围部件接口PCI直通模式部署到网络虚拟机；
    所述装置还包括：
    第一切换模块，设置为在迁移模块将所述待迁出虚拟机的SRIOV网卡资源迁移到所述待迁入虚拟机中之前，将所述待迁出虚拟机的SRIOV网卡资源从PCI直通模式切换到Macvtap直通模式。
  12. 一种控制器，包括：如权利要求9-11任一项所述的网络资源部署的装置。
  13. 一种核心网设备，包括：如权利要求12所述的控制器。
  14. 一种计算机可读存储介质，存储有计算机可执行指令，所述计算机可执行指令用于执行权利要求1-8任一项所述的网络资源部署的方法。
PCT/CN2017/080907 2016-04-18 2017-04-18 网络资源部署的方法、装置、控制器及核心网设备 WO2017181936A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610242492.4 2016-04-18
CN201610242492.4A CN107306230B (zh) 2016-04-18 2016-04-18 一种网络资源部署的方法、装置、控制器及核心网设备

Publications (1)

Publication Number Publication Date
WO2017181936A1 true WO2017181936A1 (zh) 2017-10-26

Family

ID=60115612

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/080907 WO2017181936A1 (zh) 2016-04-18 2017-04-18 网络资源部署的方法、装置、控制器及核心网设备

Country Status (2)

Country Link
CN (1) CN107306230B (zh)
WO (1) WO2017181936A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108334402A (zh) * 2018-03-07 2018-07-27 山东超越数控电子股份有限公司 一种无中心架构的虚拟化管理系统及其资源调度方法
CN109802866A (zh) * 2019-01-02 2019-05-24 郑州云海信息技术有限公司 一种网卡sr-iov功能测试方法、装置、终端及存储介质
CN114363185A (zh) * 2022-03-17 2022-04-15 阿里云计算有限公司 虚拟资源处理方法以及装置

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112130958B (zh) * 2020-09-11 2022-05-24 苏州浪潮智能科技有限公司 一种基于ovs的虚拟机热迁移方法和系统

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103201721A (zh) * 2012-08-29 2013-07-10 华为技术有限公司 虚拟机热迁移的系统和方法
CN105612722A (zh) * 2013-09-13 2016-05-25 微软技术许可有限责任公司 虚拟网络路由

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090007099A1 (en) * 2007-06-27 2009-01-01 Cummings Gregory D Migrating a virtual machine coupled to a physical device
US8321617B1 (en) * 2011-05-18 2012-11-27 Hitachi, Ltd. Method and apparatus of server I/O migration management
US9354933B2 (en) * 2011-10-31 2016-05-31 Intel Corporation Remote direct memory access adapter state migration in a virtual environment
CN102662746A (zh) * 2012-04-22 2012-09-12 复旦大学 利用虚拟机并发迁移技术降低云计算平台能源消耗的方法
CN105262604B (zh) * 2014-06-24 2019-01-08 华为技术有限公司 虚拟机迁移方法及设备

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103201721A (zh) * 2012-08-29 2013-07-10 华为技术有限公司 虚拟机热迁移的系统和方法
WO2014032233A1 (zh) * 2012-08-29 2014-03-06 华为技术有限公司 虚拟机热迁移的系统和方法
CN105612722A (zh) * 2013-09-13 2016-05-25 微软技术许可有限责任公司 虚拟网络路由

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108334402A (zh) * 2018-03-07 2018-07-27 山东超越数控电子股份有限公司 一种无中心架构的虚拟化管理系统及其资源调度方法
CN109802866A (zh) * 2019-01-02 2019-05-24 郑州云海信息技术有限公司 一种网卡sr-iov功能测试方法、装置、终端及存储介质
CN109802866B (zh) * 2019-01-02 2022-02-18 郑州云海信息技术有限公司 一种网卡sr-iov功能测试方法、装置、终端及存储介质
CN114363185A (zh) * 2022-03-17 2022-04-15 阿里云计算有限公司 虚拟资源处理方法以及装置

Also Published As

Publication number Publication date
CN107306230B (zh) 2020-12-29
CN107306230A (zh) 2017-10-31

Similar Documents

Publication Publication Date Title
US11693656B2 (en) Automatic scaling of microservices applications
US10778756B2 (en) Location of actor resources
US20230359510A1 (en) Technologies for hierarchical clustering of hardware resources in network function virtualization deployments
EP3073373B1 (en) Method for interruption affinity binding of virtual network interface card, and computer device
EP3606008A1 (en) Method and device for realizing resource scheduling
US9329889B2 (en) Rapid creation and reconfiguration of virtual machines on hosts
US9294407B2 (en) Network device load balancing in a virtualized computing environment
WO2017181936A1 (zh) 网络资源部署的方法、装置、控制器及核心网设备
US10095536B2 (en) Migration of virtual machines with shared memory
US9354905B2 (en) Migration of port profile associated with a target virtual machine to be migrated in blade servers
US9454395B2 (en) Traffic-aware data center VM placement considering job dynamic and server heterogeneity
EP3276490A1 (en) Extension of a private cloud end-point group to a public cloud
Callegati et al. Live migration of virtualized edge networks: Analytical modeling and performance evaluation
US10938688B2 (en) Network costs for hyper-converged infrastructures
EP3588856B1 (en) Technologies for hot-swapping a legacy appliance with a network functions virtualization appliance
US9537780B2 (en) Quality of service agreement and service level agreement enforcement in a cloud computing environment
US11301020B2 (en) Data center power management
JP6144639B2 (ja) ネットワーク制御装置、通信システム、ネットワーク制御方法、および、ネットワーク制御プログラム
US9612877B1 (en) High performance computing in a virtualized environment
EP3352417A1 (en) System recovery using wol
Kashyap et al. Virtual machine migration for back-end mashup application deployed on OpenStack environment
EP3126996A1 (en) Virtual-machine placement based on information from multiple data centers
JP2015156585A (ja) ネットワーク制御装置、通信システム、ネットワーク制御方法、および、ネットワーク制御プログラム
Guan et al. A novel energy efficient platform based model to enable mobile cloud applications
Prakash et al. Performance analysis of Transport Protocol during live migration of Virtual Machines‖

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17785420

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 17785420

Country of ref document: EP

Kind code of ref document: A1