WO2022233297A1 - 云网络、用于云网络的测量系统、方法、设备及存储介质 - Google Patents

云网络、用于云网络的测量系统、方法、设备及存储介质 Download PDF

Info

Publication number
WO2022233297A1
WO2022233297A1 PCT/CN2022/090914 CN2022090914W WO2022233297A1 WO 2022233297 A1 WO2022233297 A1 WO 2022233297A1 CN 2022090914 W CN2022090914 W CN 2022090914W WO 2022233297 A1 WO2022233297 A1 WO 2022233297A1
Authority
WO
WIPO (PCT)
Prior art keywords
measurement
network
network element
path
tenant
Prior art date
Application number
PCT/CN2022/090914
Other languages
English (en)
French (fr)
Inventor
祝顺民
吕彪
卢建元
康达祥
肖雄
王蕾
Original Assignee
阿里云计算有限公司
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 阿里云计算有限公司 filed Critical 阿里云计算有限公司
Publication of WO2022233297A1 publication Critical patent/WO2022233297A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/50—Testing arrangements
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters

Definitions

  • the present application relates to the technical field of cloud computing, and in particular, to a cloud network, a measurement system, method, device and storage medium for a cloud network.
  • Cloud Network is a complex network that integrates physical network and virtual network.
  • the virtual network includes multi-tenant network. It carries tenant applications and the connection and communication between applications. The application requests of tenants will pass through the multi-layer network. After processing and transmission, it reaches the destination.
  • the network quality of the cloud network has a direct impact on tenant applications.
  • the existing quality detection technologies for physical networks are not suitable for cloud networks. In this case, how to simply and accurately measure the network quality of the cloud network is very important.
  • Various aspects of the present application provide a cloud network, a measurement system, method, device and storage medium for the cloud network, so as to simply and accurately measure the network quality of the cloud network.
  • An embodiment of the present application provides a cloud network, including: a physical network and a virtual network borne on the physical network; the virtual network includes a multi-tenant network, and is responsible for connecting different end devices in the multi-tenant network
  • a network element device for traffic forwarding and interconnection the cloud network further includes: a scheduling device and an analysis device; the scheduling device is used to perceive the measurement intention of the target tenant, generate measurement rules adapted to the measurement intention, and download Sent to the source network element device, the measurement rule includes the source end device and the destination end device on the path to be tested, the source network element device is the network element device on the path to be tested; the source network element device, is used to generate a measurement request message according to the measurement rule, and forward the measurement request message, and the measurement request message is used for the source network element device and the path to be tested to receive the measurement request message.
  • the other network element equipment of the measurement request message generates measurement record information; the analysis equipment is configured to perform network quality analysis according to the measurement record information generated by the source network element equipment and
  • Embodiments of the present application further provide a measurement system for a cloud network, including: a scheduling subsystem, at least one measurement execution subsystem, and a measurement analysis subsystem; the scheduling subsystem is used to sense the measurement of a target tenant in the cloud network Intent, generate a measurement rule adapted to the measurement intention and deliver it to the target measurement execution subsystem, where the measurement rule includes the source-end device and the destination-end device on the path to be measured; the target measurement execution subsystem and all corresponding to the path to be measured; the target measurement execution subsystem is configured to generate a measurement request message according to the measurement rule, and inject the measurement request message into the path to be measured, so that the At least part of the network element devices on the path forward the measurement request message and generate measurement record information; the measurement analysis subsystem is configured to perform network quality analysis according to the measurement record information generated by the at least part of the network element devices.
  • An embodiment of the present application further provides a network quality measurement method, including: sensing a measurement intention of a target user in a cloud network, and generating a measurement rule adapted to the measurement intention, where the measurement rule includes a source end on a path to be measured device and destination device; generate a measurement request message according to the measurement rule, and inject the measurement request message into the path to be tested, so that at least some network element devices on the path to be tested generate measurement record information; perform network quality analysis according to the measurement record information generated by the at least part of the network element devices.
  • An embodiment of the present application further provides a cloud computing device, including: a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program, so as to implement the method of the present application steps in the example.
  • the embodiments of the present application further provide a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the processor can implement the steps in the method embodiments of the present application.
  • Embodiments of the present application further provide a computer program product, including computer programs/instructions, which, when executed by a processor, enable the processor to implement the steps in the method embodiments of the present application.
  • a measurement rule is generated according to the measurement intention, and based on the measurement rule, the method of bypassing packets is injected into the network element equipment on the path to be measured.
  • the measurement request message is used to analyze the network quality by means of the measurement record information generated when the measurement request message passes through different network element devices, so that the network quality of the cloud network can be measured simply, effectively and accurately.
  • the measurement process can be greatly simplified by the intent of tenant measurement requests, and tenants are not aware of the measurement process, and do not need to manage complex measurement rules, which is beneficial to improve tenant experience; method, there is no need to deploy measurement services in the tenant's network environment, which is beneficial to reduce the intrusion into the tenant's network environment; at the same time, since the embodiment of the present application is to actively inject packets, the dependence on the actual application traffic of the tenant can be avoided, even if the tenant does not Network quality can also be measured while generating actual application traffic.
  • FIG. 1a is a schematic structural diagram of a measurement system for a cloud network provided by an exemplary embodiment of the present application
  • FIG. 1b is a schematic diagram of a process of injecting a measurement request message into a path to be tested to generate measurement record information according to an exemplary embodiment of the present application;
  • FIG. 1c is a schematic diagram of the internal structure of a measurement and analysis subsystem provided by an exemplary embodiment of the present application.
  • FIG. 1d is a schematic diagram of the use state of the measurement and analysis subsystem provided by the exemplary embodiment of the present application when it is deployed independently of the cloud network;
  • FIG. 2a is a schematic structural diagram of a cloud network provided by an exemplary embodiment of the present application.
  • FIG. 2b is a schematic structural diagram of another cloud network provided by an exemplary embodiment of the present application.
  • FIG. 3a is a schematic flowchart of a network quality measurement method provided by an exemplary embodiment of the present application.
  • 3b is a schematic flowchart of a method for generating a measurement rule provided by an exemplary embodiment of the present application
  • 3c is a schematic flowchart of a message transmission method provided by an exemplary embodiment of the present application.
  • FIG. 4 is a schematic structural diagram of an apparatus for measuring network quality provided by an exemplary embodiment of the present application.
  • FIG. 5 is a schematic structural diagram of a cloud computing device according to an exemplary embodiment of the present application.
  • a measurement system is provided, and the measurement system can be used to measure the network quality of the cloud network, but is not limited to this .
  • the measurement system can also be applied to a physical network to measure network quality for the physical network.
  • the measurement system can perceive the measurement intention of the tenant in the cloud network, generate measurement rules according to the measurement intention, and use the bypass packet injection method to send packets on the path to be measured based on the measurement rules.
  • the network element equipment injects measurement request packets, and then uses the measurement record information generated when the measurement request packets pass through different network element equipment to perform network quality analysis, which can simply, effectively and accurately measure the network quality of the cloud network.
  • the application of the measurement system to the cloud network is taken as an example for description.
  • cloud networks are complex networks that integrate physical networks and virtual networks, and virtual networks include multi-tenant networks, that is, cloud networks include physical networks and virtual networks.
  • the physical network includes physical machines such as servers, cabinets, routers, and switches, as well as physical connecting lines, such as coaxial cables, network cables, and optical fibers, used to implement network connections between these physical machines.
  • the virtual network is carried on the physical network and is a logical network implemented based on the virtualization technology on the basis of the physical network. It can virtualize the physical network resources, so that the physical network resources can be upgraded to virtualized and dynamically allocated virtual resources.
  • the virtual network includes a multi-tenant network, and the network environments of different tenants are isolated from each other.
  • Each tenant network includes but is not limited to: all data that can be identified as the tenant in the cloud network, such as in the cloud Accounts and statistical information (Accounting Data) created for the tenant in the network, as well as various virtualized resources such as various data set for the tenant in the cloud network and virtualized end devices configured by the tenant.
  • the end device in the tenant network refers to a virtualized device that can initiate or terminate network traffic. Generally, the end device can provide a completely isolated network environment for the tenant, and is responsible for carrying the tenant's applications.
  • the sending and receiving of network traffic can be implemented in but not limited to: virtual machines (VMs), containers, elastic instances (ECSs), and configuration objects for subsequent network element devices (such as virtual switches). , cloud database, etc.
  • VMs virtual machines
  • ECSs elastic instances
  • configuration objects for subsequent network element devices such as virtual switches
  • cloud database etc.
  • the tenant generally refers to the user who uses the cloud network or various resources in the cloud network.
  • the implementation form of the tenant network is not limited.
  • it can be implemented as a virtual private cloud (Virtual Private Cloud, VPC).
  • VPC Virtual Private Cloud
  • tenants their own VMs, cloud databases and other terminals can be deployed in their VPC. equipment.
  • the virtual network also includes network element devices that implement traffic forwarding and network interconnection between different end devices in the multi-tenant network.
  • the network element device is mainly used to forward traffic between different end devices, process packets, and implement network interconnection between different end devices, such as a virtual switch, a virtual gateway, and the like.
  • an overlay network may be implemented in a virtual network by using an overlay technology, the overlay network is between the physical network and the tenant network, and the above network element devices may be implemented in the overlay network.
  • the overlay network is a bridge between the tenant network and the physical network, which enables the interconnection between the mid-end devices of the tenant network to escape the constraints of the physical network, and provides conditions for the final realization of flexible definition, on-demand allocation, and on-demand adjustment of network resources.
  • the entire cloud network can be implemented as a three-layer network architecture including a physical network, an overlay network, and a multi-tenant network, but the division of the cloud network architecture is not limited to this. No matter what kind of network architecture it is, by packaging the configuration information of various resources in the virtual network and finally presenting it to the tenant in the form of a configuration object, the tenant can configure the configuration object according to their own needs, so as to obtain end devices that meet their own needs. , subnets and other resources.
  • cloud networks From the above introduction to cloud networks, it can be seen that network quality detection of cloud networks is no longer limited to physical machines or between physical machines, but needs to go deep into virtualized end devices (such as virtual machines).
  • a device is a logical node, and its deployment and topology are quite different from that of a physical machine.
  • the topology of the tenant network In the cloud network, compared with the topology of the physical network, the topology of the tenant network is flexible and changeable. According to the different needs of the tenant, these virtualized end devices (such as virtual machines) can be flexibly increased or decreased.
  • the interconnection relationship between end devices and between end devices and other networks eg, private networks
  • the tenant directly detects the network quality of the cloud network, the tenant needs to perform complex measurement work, including deploying measurement services, managing measurement objects and measurement rules, etc., and these tasks also need to adapt to changes in the virtual network topology. , which is a heavy workload for tenants, and it is difficult to measure network quality.
  • the measurement system provided by the embodiment of the present application can solve the technical problem that the network quality measurement of the cloud network is difficult.
  • the measurement system 100 provided by the embodiment of the present application that can be used for a cloud network includes: a scheduling subsystem 101, at least one measurement execution subsystem 102, and a measurement Analysis subsystem 103 .
  • each measurement execution subsystem 102 is connected in communication with the scheduling subsystem 101 and the measurement analysis subsystem 103 respectively.
  • the scheduling subsystem 101 , the measurement execution subsystem 102 , and the measurement analysis subsystem 103 may be implemented in software.
  • the scheduling subsystem 101 , the measurement execution subsystem 102 , and the measurement analysis subsystem may be deployed.
  • the respective software codes of the systems 103 In this embodiment, the deployment locations of the scheduling subsystem 101 , the measurement execution subsystem 102 , and the measurement analysis subsystem 103 are not limited, and they can be deployed at appropriate network locations according to application requirements.
  • the scheduling subsystem 101 is used to sense the measurement intent of the target tenant in the cloud network on the one hand, and generate measurement rules adapted to the measurement intent;
  • the measurement rules are issued to the appropriate measurement execution subsystem 102, so that the measurement execution subsystem 102 performs operations related to network quality measurement according to the measurement rules and generates measurement record information ;
  • the network quality analysis is performed by the measurement analysis subsystem 103 according to the measurement record information, and a network quality analysis result is generated. Further optionally, as shown in FIG.
  • the network quality analysis result may also be fed back to the scheduling subsystem 101, which further affects the process of perceiving the measurement intention of the target tenant by the measurement execution subsystem 102.
  • the scheduling subsystem 101 is responsible for sensing the measurement intention of the target tenant and generating measurement rules adapted to it according to the measurement intention, which can liberate the target tenant from network quality measurement, and the target tenant does not need to maintain and manage measurement rules, which is conducive to improving The target tenant's experience in using the cloud network.
  • the target tenant refers to the tenants in the cloud network, the number of which may be one or more, may be all tenants in the cloud network, or may be part of the tenants in the cloud network, which may be determined by network quality measurement requirements. For example, if you need to know the network quality of the entire cloud network from the perspective of the cloud network provider, you can use all tenants in the cloud network as target tenants, perceive the measurement intention of each tenant, and measure the network quality for each tenant. Measure to get the network quality of the entire cloud network.
  • these tenants can be used as target users to perceive the measurement intentions of these tenants, and perform network quality measurement for these tenants, so as to meet the network quality measurement requirements of these tenants.
  • you need to know the network quality of a certain area covered by the cloud network from the perspective of the area you can take the tenants in the area as the target tenants, perceive the measurement intentions of these tenants, and measure the network quality of these tenants. measurement to obtain the network quality in the area.
  • the scheduling subsystem 101 perceives the measurement intention of the target tenant; the measurement intention of the target tenant is the intentionalization of the measurement request of the target tenant, which can reflect what kind of measurement request the target tenant needs. Closer to the user's language expression or description. It can be understood that the measurement intent is an expression of the measurement request of the target tenant in a language close to the user or in a description manner.
  • the measurement request here refers to a request to measure the network quality of the cloud network.
  • the scheduling subsystem 101 can sense the measurement intention of the target tenant in various ways, which is not limited.
  • the scheduling subsystem 101 can monitor whether the network topology of the target tenant changes; in the case where the network topology of the target tenant does not change, the measurement intention of the target tenant generally does not change;
  • the network topology changes for example, the target tenant purchases a new virtual machine, deletes the existing virtual machine, or changes the configuration information of the virtual machine, etc.
  • the measurement intention of the network quality usually changes;
  • the subsystem 101 can perceive the measurement intention of the target tenant according to the network topology change information of the target tenant.
  • the scheduling subsystem 101 may obtain application demand change information submitted by the target tenant, generate network topology change information of the target tenant according to the application demand change information submitted by the target tenant, and determine the target tenant's network topology change information according to the network topology change information of the target tenant.
  • Measurement intent changes in the application requirements of the target tenant will directly affect the network topology of the target tenant. For example, due to the increase in the number of downstream users, the target tenant needs to add a virtual machine, and the network topology of the target tenant will change due to the addition of a virtual machine.
  • This virtual machine may need to be interconnected with a load balancer, or the A virtual machine needs to be connected to a dedicated line, or the virtual machine needs to be interconnected with other existing virtual machines, which will change the target tenant's intention to measure the network quality.
  • the implementation manner in which the scheduling subsystem 101 obtains the application demand change information submitted by the target tenant is not limited. The following example illustrates:
  • the scheduling subsystem 101 can be implemented as a portal for the target tenant to interact with the cloud network provider, and the application requirement information submitted by the target tenant at the beginning or the application requirement change information submitted subsequently can be passed through the scheduling sub-system.
  • the system 101 makes a submission so that the cloud network provider can provide the corresponding cloud computing service for the target tenant.
  • the scheduling subsystem 101 can directly receive the application requirement change information submitted by the target tenant.
  • the scheduling subsystem 101 may provide a web page, application page or command window for the target tenant, and the target tenant opens the web page, application page or command window provided by the scheduling subsystem 101 on the terminal device used by the target tenant, The web page, application page or command window submits application requirement change information to the scheduling subsystem 101 .
  • the virtual network of the cloud network further includes a network controller, where the network controller is a portal for interaction between the tenant and the cloud network provider.
  • the network controller may be Some distributed software, for example, can be an API.
  • the target tenant can submit application requirement information or application requirement change information to the cloud network provider.
  • the network controller may provide a web page, application page or command window for the target tenant, and the target tenant opens the web page, application page or command window provided by the network controller on the terminal device used by the target tenant, and the target tenant opens the web page, application page or command window provided by the network controller, , application page or command window to submit application requirement change information to the network controller.
  • the scheduling subsystem 101 communicates with the network controller in the cloud network, and acquires the application requirement change information submitted by the target tenant from the network controller.
  • the scheduling subsystem 101 may be hung next to the network controller, and the application demand change information submitted by the target tenant is shunted from the network controller to the scheduling subsystem 101 by means of bypass collection. middle.
  • the network controller may provide a subscription service for the scheduling subsystem 101, and the scheduling subsystem 101 may be registered as a subscriber of the network controller for subscribing to the network controller for the application requirement change information submitted by the tenant, In this way, when the network controller receives the application requirement change information submitted by any tenant, it can actively provide the application requirement change information submitted by the tenant to the scheduling subsystem 101 .
  • the scheduling subsystem 101 can generate the target tenant's network according to the application demand change information and in combination with the maintained network topology of the cloud network. topology change information; and then determine the measurement intention of the target tenant according to the network topology change information.
  • the scheduling subsystem 101 can add a new virtual machine and a network connection related to the new virtual machine on the basis of the original network topology of the target tenant to obtain the network topology of the target tenant change information; further, sensing that the target tenant may need to measure the network quality between the existing virtual machine and the newly added virtual machine, the measurement intent can be expressed as: measuring the difference between the existing virtual machine Vm1 and the newly added virtual machine Vm2 the network quality in between.
  • the scheduling subsystem 101 may increase the connection between the existing virtual machine and the public network on the basis of the original network topology of the target tenant.
  • the measurement intent can be expressed as: measure the network quality of public network-12345 (measure network-12345).
  • the scheduling subsystem 101 can generate measurement rules adapted to the measurement intention.
  • the measurement rule includes relevant information required to perform network quality measurement on the cloud network according to the network quality measurement requirements.
  • the information includes at least information that can indicate which paths between end devices need to perform network quality measurement. For example, it may include but Not limited to: the identification information of the source end device and the destination end device, the source end device and the destination end device can uniquely determine a network path. Measurement path.
  • the identification information here may be any information that can uniquely identify the source end device or the destination end device, for example, may be an IP address, a MAC address, or a device serial number.
  • the source device and the destination device may be one or more groups, depending on the network quality measurement requirements.
  • the measurement rules may also include the communication protocols used for network communication between the source end device and the destination end device, such as UDP protocol, TCP protocol, etc., so as to facilitate the use of communication supported by the source end device and the destination end device.
  • the protocol performs network quality measurements on the path between the two.
  • the measurement rule generated according to the measurement intention can be expressed as: source device (Src): Vm1, destination device ( Dst): Vm2, communication protocol (protocol): UDP.
  • the scheduling subsystem 101 may determine the source end device and the destination end device on the path to be measured according to the measurement intention of the target tenant and in combination with the network configuration information of the target tenant; The source device and the destination device generate measurement rules.
  • the network configuration information of the target tenant refers to various information configured by the target tenant through the tenant network layer, for example, including but not limited to: routing table entry information, access control list (ACL) information, network element device information, network topology information, etc.
  • the source-end device to be measured can be parsed from the measurement intent of the target tenant, for example, it can be a newly added virtual machine of the target tenant, or a public network or private line that needs to be accessed
  • determine the network topology where the source device is located from the topology of the entire cloud network and obtain potential paths that have an access relationship with the source device according to the network topology where the source device is located and combined with the network configuration information of the target tenant;
  • the potential path can be one or more; the end device at one end of the potential path is the source end device, and the destination end device corresponding to the source end device is determined from the other end devices on the potential path.
  • the implementation manner of determining the destination end device corresponding to the source end device from the other end device on the potential path is not limited. The following example illustrates:
  • the other end devices on all potential paths may be regarded as the destination end devices corresponding to the source end devices by default, which means that all potential paths will be regarded as Measurements at least give insight into the network quality of all potential paths.
  • the maximum number of paths allowed for each measurement can be set, denoted as N, where N is a fixed value, for example, it can be an integer value such as 1, 2, 3, or 4, then in the potential path
  • N is a fixed value, for example, it can be an integer value such as 1, 2, 3, or 4, then in the potential path
  • at least one potential path with a number not greater than N can be randomly selected from it, and the other end device on the selected at least one potential path is used as the destination end device corresponding to the source end device.
  • selected At least one potential path of the selected path will be regarded as the path to be tested, and the network quality of at least one selected potential path can be learned through measurement.
  • the target tenant is allowed to select the measurement method to be used, and the user-selectable measurement method includes, but is not limited to: a full coverage measurement method and a selective measurement method.
  • the full coverage measurement method refers to a method of measuring all existing potential paths
  • the selective measurement method refers to a method of selectively measuring some potential paths.
  • the target user can choose the measurement method to be used; if the target tenant chooses the full coverage measurement method, the other end devices on all potential paths can be used as the destination end device corresponding to the source end device, That is, all potential paths will be regarded as the paths to be tested, and at least the network quality of all potential paths can be known through measurement; if the target tenant chooses the selective measurement method, some potential paths can be selected from all potential paths, and the selected The other end device on the part of the potential path is regarded as the destination end device corresponding to the source end device, that is, the selected part of the potential path will be regarded as the path to be tested, and the network quality of at least the selected part of the potential path can be known through measurement. .
  • this embodiment of the present application does not limit this.
  • some potential paths may be randomly selected, or, some potential paths may be selected according to the maximum number N of paths allowed for each measurement, or, It is also possible to allow the target tenant to set selection conditions, such as the type of the device at the other end, regional location, network access time, etc., and select some potential paths according to the selection conditions set by the target tenant.
  • the scheduling subsystem 101 can display a human-computer interaction interface to the target tenant, and a measurement method selection item is set on the human-computer interaction interface, and the selection item includes at least a full coverage measurement method and selection.
  • the measurement method selected by the target tenant is determined; in response to the selection operation initiated by the target tenant on the measurement method option, the measurement method selected by the target tenant is determined, wherein the measurement method selected by the target tenant or the full coverage measurement method, Or a selective measure.
  • the scheduling subsystem 101 may also obtain the measurement method selected and used by the target tenant through the network controller in the cloud network.
  • the network controller may display a human-computer interaction interface to the target tenant, and a measurement method selection item is set on the human-computer interaction interface, and the selection item includes at least a full-coverage measurement method and a selective measurement method for the target tenant to perform.
  • the implementation form of the human-computer interaction interface is not limited, for example, it may be a web page, an application page, or a command window.
  • the scheduling subsystem 101 may deliver the measurement rules to the measurement execution subsystem 102, and the measurement execution subsystem 102 performs operations related to network quality measurement according to the measurement rules.
  • the measurement execution subsystem 102 may be one or multiple. Regardless of whether there are one or more measurement execution subsystems 102, this embodiment does not limit the deployment location of the measurement execution subsystem 102.
  • the measurement execution subsystem 102 may be deployed at any location on a path except for the end device, that is, in this In the embodiment, the measurement execution subsystem 102 is not deployed in the tenant's end device, which can reduce the intrusion of the network quality measurement to the tenant's network environment, so that the tenant does not need to manage and maintain complex measurement rules.
  • the scheduling subsystem 101 will directly issue the measurement rules to the measurement execution subsystem 102, and the measurement execution subsystem 102 Perform operations related to network quality measurement for the path under test.
  • the scheduling subsystem 101 in this embodiment is also responsible for scheduling the measurement execution subsystem 102, so that the measurement execution subsystem 102 adapted to the path to be tested is used for network execution. quality measurement.
  • the multiple measurement execution subsystems 102 may be deployed in a distributed deployment manner, that is, the multiple measurement execution subsystems 102 will be deployed in different locations.
  • the physical resources occupied by the measurement execution subsystem 102 may also vary.
  • the scheduling subsystem 101 may preferentially select the measurement execution subsystem that is closer to the path to be measured, and/or select the measurement execution subsystem 102 with better execution performance, and /or, select the measurement execution subsystem 102 with less load, etc., which are not limited.
  • the scheduling subsystem 101 can issue the above measurement rules to the target measurement execution subsystem, the target measurement execution subsystem is the measurement execution subsystem to be scheduled in at least one measurement execution subsystem 102, and the target measurement execution subsystem is The measurement execution subsystem adapted to or corresponding to the path to be tested may, for example, be the closest to the path to be tested, or be deployed on the path to be tested, or have better execution performance.
  • the target measurement execution subsystem is specifically used to: when it is scheduled, according to the measurement rules issued by the scheduling subsystem 101, generate a measurement request message in the name of the source device, and inject the measurement request message into the under-measured on the path, so that at least some network element devices on the path to be tested forward the measurement request message and generate measurement record information.
  • the target measurement execution subsystem injects the measurement request message into the path to be tested, specifically, injecting the measurement request message into a certain network element device on the path to be tested, so that the measurement request message can be sent to the path to be tested.
  • the message is continuously forwarded from the network element device toward the destination device.
  • the network element device on which the measurement request message is injected on the path to be measured is referred to as the source network element device, and the source network element device may be any network element device on the path to be measured, and the specific network element device can be determined according to the measurement requirements. Depends.
  • the measurement request message refers to the message sent to the destination end device in the name of the source end device. It should be noted that the measurement request message is not generated by the source end device, nor is it sent from the source end device, but It is generated by the target measurement execution subsystem instead of the source device and injected into the path to be tested, as shown in Figure 1b.
  • the message format of the measurement request message is not limited, and any message format that can reflect "sent to the destination device in the name of the source device" is applicable to the embodiment of the present application.
  • the following is a standardized format of a measurement request message given by an embodiment of the present application, which can support various communication protocols, such as TCP or UDP protocols.
  • the measurement request message includes at least: a protocol header and a measurement header; wherein, the protocol header can be an IP header or a UDP header, etc., to support different communication protocols; about each header includes
  • the field information is as follows:
  • dscp the combination of IP precedence and service type fields, its value can be but not limited to: 63;
  • sip IP address of the source device
  • dport for example, it can be 5000, but not limited to this;
  • tenant-id ID of the tenant to which the source device belongs, which can represent the virtual network topology of the tenant;
  • sip IP address of the source device
  • mea_src_tenant_id ID of the network space to which the source NE device belongs.
  • a tenant can have multiple network spaces, and different network spaces can be distinguished by the ID of the network space;
  • mea src_ip the IP address of the source network element device and/or the IP address of the physical machine that carries the source network element device;
  • timestamp the timestamp when the measurement request packet is sent by the source network element device, which is regarded as the first timestamp
  • mea_reply_tenant_id used to carry the ID of the tenant of the NE that returns the measurement reply message in the measurement reply message, reserved in the measurement request message, and the default value is 0;
  • mea reply_ip used to carry the IP address of the network element device that returns the measurement reply message in the measurement reply message, and is reserved in the measurement request message;
  • reply-timestamp It is used to carry the timestamp of the returned measurement reply packet in the measurement reply packet, and is reserved for the measurement request packet.
  • the target measurement execution subsystem when it generates the measurement request message, it can specifically generate a protocol header according to the information of the source device and the destination device.
  • the IP address of the device generates an IP header; if it is a UDP header, the UDP header is generated according to the port number and protocol type of the source device and the destination device; furthermore, according to the ID of the target tenant (corresponding to the tenant-id field) , the detailed information of the source network element device, and the first timestamp (corresponding to the timestamp field) of the measurement request packet, to generate the measurement header of the measurement request packet; according to the protocol header and the measurement header, the measurement request packet is generated.
  • the detailed information of the source network element device includes but is not limited to: the ID of the network space to which the source network element device belongs (corresponding to the mea_src_tenant_id field), the IP address of the source network element device, and/or the IP address of the physical machine that bears the source network element device (corresponding to the mea src_ip field).
  • the target measurement execution subsystem if it is not deployed in the source network element device, the time when it injects the measurement request message into the source network element device can be used as the first timestamp; if it is deployed in the source network element device. In the network element device, the time when the source network element device sends the measurement request message can be known as the first timestamp.
  • the above-mentioned source network element device refers to the network element device to which the measurement request message on the path to be tested is injected.
  • the manner in which the target measurement execution subsystem injects the measurement request message into the source network element device will be different.
  • a measurement execution subsystem may be installed next to each network element device in the cloud network. Based on this, after the target measurement execution subsystem generates a measurement request message, it may send the measurement request message to Its corresponding target measurement execution subsystem.
  • the measurement execution subsystem may be directly deployed in each network element device in the cloud network.
  • the process of injecting the measurement request message into the source network element device by the target measurement execution subsystem is actually The above is the process of generating the measurement request message by the target measurement execution subsystem deployed on the source network element device.
  • the generation time of the measurement request message is basically the same as or the same as the sending time of the measurement request message.
  • the source network element device is not specifically limited, and the source network element device may be any network element device on the path to be measured.
  • the path to be tested includes not only the source device and the destination device, but also at least one network element device connected between the source device and the destination device, and the target measurement execution subsystem can
  • the request message is sent to any network element device on the path to be tested, that is, the source network element device.
  • the source network element device forwards the measurement request message, so that the measurement request message continues to be sent to other networks on the path to be tested. meta-device for forwarding.
  • the specific network element device to which the measurement request message is injected and forwarded to can be determined according to the network quality measurement requirements.
  • the network element device to which the measurement request message is finally forwarded is called the target network element device.
  • a measurement request message may be injected into the starting network element device of the partial path, and the starting network element device will forward the measurement request message. , the measurement request message continues to be forwarded by the subsequent network element device until the end network element device of the partial path.
  • the source network element device above is the starting network element device, and correspondingly, the target network element device is the ending network element device.
  • the local path refers to a section of the path to be tested.
  • the target measurement execution subsystem may limit the forwarding times of the measurement request message, and determine by the forwarding times that the measurement request message will not be forwarded continuously after reaching the end network element device.
  • the target measurement execution subsystem may also send a termination instruction to the ending network element device to instruct the ending network element device not to continue to forward the measurement request message to the outside.
  • the target measurement execution subsystem may send a measurement request message to a first network element device directly connected to the source device on the path to be measured, and the first network element device sends the measurement request message Forwarded, so that the measurement request message continues to be forwarded on the path to be tested to the second network element device, where the second network element device is a network element device directly connected to the destination device on the path to be tested.
  • the measurement request message is injected into the first network element device and is gradually forwarded to the second network element device as an example for illustration.
  • the source network element device above is the first network element device
  • the target network element device is the second network element device.
  • each intermediate network element device on the path to be tested will forward the measurement request message, and can generate corresponding measurement record information, which can not only measure the network quality of the entire path, but also measure the network quality of the intermediate path.
  • the network quality of each network element device is measured, and it has the ability to quickly and accurately locate network problems.
  • the second network element device after receiving the measurement request message, the second network element device does not forward the measurement request message, but discards the measurement request message after generating the measurement record information, In this way, the intrusion of the measurement request packet into the destination device can be reduced, so that the tenant is unaware of the network measurement.
  • the second network element device discards the measurement request message and does not forward it to the destination device as an example for illustration.
  • the second network element device can also forward the measurement request packet to the destination end device, and the destination device generates measurement record information in the same way as the previous network element device.
  • whether the destination device needs to perform network quality measurement can be determined by the tenant to which the destination device belongs.
  • the tenant can choose to enable the network quality measurement function of the destination device, and the second network element device can continue.
  • the measurement request message is forwarded to the destination device; if the tenant does not enable the network quality measurement function of the destination device, the second network element device will directly discard the measurement request message and will not continue to forward it.
  • the message format of the measurement reply message is not limited. Similar to the measurement request message, the embodiment of the present application provides a standardized format of the measurement reply message. As described below, the measurement reply message at least Including: protocol header and measurement header; wherein, the protocol header can be an IP header or a UDP header, etc., to support different communication protocols, and the field information included in each header is as follows:
  • dscp The combination of IP precedence and service type fields, the value of which can be but not limited to: 62
  • sip IP address of the destination device
  • dport destination port, the same as the measurement request message, its value can directly copy the value of the UDP header in the measurement request message;
  • the measurement header in the measurement reply message contains the same field information as the measurement header in the measurement request message. You can directly copy the field values in the measurement header in the measurement request message, but add mea_reply_tenant_id , mea_reply_ip, and reply-timestamp are used to measure the field values of reply packets. Based on this, it can be known that the field information contained in the measurement header in the measurement reply message is as follows:
  • uuid The unique ID of the measurement reply message, which can be the same as or corresponding to the value in the measurement request message;
  • tenant-id ID of the tenant to which the source device belongs
  • sip IP address of the source device
  • mea_src_tenant_id ID of the network space to which the source NE device belongs;
  • mea src_ip the IP address of the source network element device and/or the IP address of the physical machine that carries the source network element device;
  • timestamp the timestamp when the measurement request packet was sent, that is, the first timestamp
  • mea_reply_tenant_id Returns the ID of the tenant to which the NE that measures the reply message belongs to;
  • mea reply_ip Returns the IP address of the network element device that measures the reply message
  • reply-timestamp Returns the timestamp of the reply measurement packet.
  • the protocol header in the measurement reply message can be generated according to the information of the source device and the destination device; according to the network element itself
  • the ID of the tenant the detailed information of the network element itself, the first timestamp and other information are used to generate the measurement header in the measurement response packet; the measurement response packet is generated according to the protocol header and the measurement header.
  • the network element device when the network element device generates the measurement response message, some information can be directly copied from the measurement request message; reply message.
  • the source network element device in addition to forwarding the measurement request message, it can also receive the measurement response message returned by other network element devices.
  • the function of the measurement request message is to allow the network element device or end device through which the measurement request message passes to generate measurement record information.
  • the measurement record information is some information related to network quality measurement recorded by the network element equipment or end equipment through which the measurement request message passes, such as tenant information, network element equipment information, and the sending and receiving time of the measurement request message and other information.
  • the tenant information can be used to analyze which tenants are involved in the path to be tested and whether it is cross-tenant;
  • the information of the network element equipment can be used to analyze which network element equipment the measurement request packet actually passes through, and further combine the network element equipment that the measurement request packet should pass through.
  • the element device can analyze the packet loss of the path to be tested, the tenant network or the entire cloud network; measuring the sending and receiving time of request packets can be used to analyze the network delay inside the network element device, the network delay of the tenant network or the entire path, Furthermore, the network latency of the tenant network or the entire cloud network can be analyzed.
  • the measurement record information is divided into two types, one is path record information, which is mainly used to record the path information passed by the measurement request packet; the other is delay record information, which is mainly used for Record the transmission delay of the measurement request message during the transfer process.
  • the path record information can be generated for each network element device that sends the measurement request message; the delay record information can be generated by the source network element device, that is, the above-mentioned starting network element device or the first network element device, and can be generated according to the measurement request.
  • the information related to the message and the measurement reply message generates the delay record information of the measurement request message.
  • the information formats of the path record information and the delay record information are not limited.
  • An exemplary format of the path record information and the delay record information provided in this embodiment is as follows:
  • Tenant-id The ID of the tenant to which the NE device that generates the path record information belongs.
  • the tenant ID may be different from the tenant ID in the measurement request packet;
  • IP address of the network element device that generates the path record information and/or the IP address of the physical machine that carries the network element device; the IP address of the network element device may be the same as the IP address of the physical machine, or it may be Are not the same;
  • GEN-POINT reserved field, which is not limited
  • GEN-TS records the timestamp of the generated path record information, which is called the second timestamp and/or the third timestamp for the convenience of distinction. See below for the explanation of the second timestamp or the third timestamp;
  • GEN-DROP-CODE If the packet is discarded, fill in the reason for the packet loss.
  • the reason for the packet loss can be hardware reasons, speed limit reasons, etc.;
  • a target network element device a boolean variable whose value is yes or no;
  • Next-hop NE IP IP address of the next-hop NE device
  • Measurement header It is the same as the measurement header in the measurement request message, and its specific value can directly copy the value of the measurement header in the measurement request message.
  • the fields included in the above path record information can be flexibly adjusted according to application requirements, and can be increased or decreased, and are not limited to these fields.
  • the source network element device ie the first network element device or the originating network element device
  • other network element devices that receive the measurement request message can record the ID of the tenant to which the network element itself belongs, the detailed information of the network element itself, At least one information of the second timestamp of the measurement request packet and the third timestamp of the forwarding measurement request packet is received to generate path record information.
  • other network element devices that receive the measurement request message may generate a path record information
  • the GEN-TS field may include two timestamps, namely the second timestamp and the third timestamp;
  • the second time stamp represents the time when the measurement request message is received
  • the third time stamp represents the time when the measurement request message is forwarded.
  • other network element devices that receive the measurement request message may generate two path record information, that is, when receiving the measurement request message, generate one path record information, then the path record information
  • the GEN-TS field in the information records the second time stamp; when forwarding the measurement request message, a path record information is generated again, and the GEN-TS field in the path record information records the third time stamp.
  • Tenant-id The ID of the tenant to which the network element device that generates the delay record information (that is, the source network element device) belongs. When the source network element device generates delay record information, the tenant ID is the same as the tenant in the measurement request packet. same ID;
  • GEN-IP the IP address of the network element device (that is, the source network element device) that generates the delay record information and/or the IP address of the physical machine that carries the network element device; wherein, these two IP addresses may be the same, or may be the same. Are not the same;
  • GEN-POINT reserved field, which is not limited
  • GEN-TS record the timestamp of sending the measurement request message, that is, the first timestamp
  • NE IP the IP address of the NE device that generates the measurement reply message
  • Measurement header It is the same as the measurement header in the measurement reply message, and its value can be directly copied to the value of the measurement header in the measurement reply message, but the source network element device will modify the measurement header after receiving the measurement reply message.
  • reply-timesatmp field change it to the timestamp when the measurement reply request message is received, which is regarded as the fourth timestamp.
  • the source network element device may generate the delay record information according to the first time stamp of sending the measurement request message and the fourth time stamp of receiving the measurement reply message.
  • the source network element device may read the first time stamp from the measurement reply message, and record the first time stamp in the delay record information; in addition, record the time stamp of the received measurement reply message as the fourth time stamp , and record the fourth time stamp into the delay record information.
  • the above-mentioned source network element equipment and other network element equipment can report the path record information or delay record information to the measurement analysis subsystem 103; on the basis of the path record information and delay record information, the measurement analysis subsystem 103 can Network quality analysis is performed in at least one dimension of paths, tenants, and cloud networks. Specifically, the measurement and analysis subsystem 103 can analyze the network delay of the path to be tested, the packet loss rate on the path to be tested, the network delay of the target tenant, the network delay of the target tenant, and the At least one of the packet loss rate, the network latency of the entire cloud network, and the packet loss rate of the entire cloud network.
  • an internal implementation structure of the measurement analysis subsystem 103 includes: a path analysis module 103a, a packet loss statistics module 103b, a delay calculation module 103c, and a result aggregation module 103d.
  • the path record information generated by the source network element equipment or other network element equipment is sent to the path analysis module 103a, and the path analysis module 103a counts the path records on the same path to be measured according to the packet ID in the measurement header in the path record information.
  • Path record information that is, the path record information containing the same packet ID is the path record information on the same path to be tested, classify the path record information from the path dimension, and classify the path record information on different paths to be tested. They are respectively provided to the packet loss statistics module 130b.
  • the packet loss statistics module 103b can analyze the path record information on each path to be measured, combined with the tenant ID in the path record information and the IP of the network element device, and can analyze which network element devices the measurement request packet has passed through, and further combine the path record information.
  • the packet loss reasons in the information can also analyze which network element devices have lost packets and what are the reasons for the packet loss; further, based on this information, the packet loss rates on different paths to be tested can be counted from the path dimension; further, combined with The tenant ID can also count the paths to be tested corresponding to the same tenant, and according to the packet loss rate on the paths to be tested corresponding to the same tenant, the packet loss rate of the tenant's network can be counted from the tenant dimension.
  • the packet loss rate of different tenant networks is calculated to calculate the packet loss rate of the entire cloud network.
  • the packet loss statistics module 130b provides the calculated packet loss rates in each dimension to the result aggregation module 103d.
  • the delay record information generated by the source network element device is sent to the delay calculation module 103c; the delay record module 103c can count the delay on the same path to be measured according to the packet ID in the measurement header in the delay record information record information; further, according to the first timestamp and the second timestamp in the delay record information, the network delay between the source network element device and different network element devices can be measured; further, from the path dimension, The network delay of the entire path to be tested is calculated; further, the network delay of the entire cloud network can be calculated from the network delay of different paths to be tested from the cloud network dimension.
  • the delay calculation module 103c provides the calculated packet loss rate in each dimension to the result aggregation module 103d.
  • the path record information can also be sent to the delay calculation module 103c; the delay calculation module 103c can combine the message ID, network element IP and other information in the path record information to identify that the same network element device is receiving measurement.
  • Path record information generated when requesting packets and forwarding measurement request packets; according to the second and third timestamps recorded in the two path record information, calculate the internal processing delay of the NE, and calculate the The processing delay inside the network element device is provided to the result aggregation module 103d.
  • the result aggregation module 103d can summarize the packet loss rate in each dimension and the network delay in each dimension, perform network quality analysis according to the network quality measurement requirements, and output the network quality analysis result.
  • the network quality analysis result can also be output to the scheduling subsystem 101 to influence the intentional perception of the scheduling subsystem 101 to form a measurement closed-loop system and improve the network quality measurement effect.
  • the tenant can be released from the network quality measurement. It avoids the tenant's management of complex measurement rules and simplifies the network quality measurement work.
  • the bypass packet injection method in this embodiment is active packet injection, which can avoid dependence on the actual application traffic of the tenant, and can measure the network quality even when the tenant does not generate actual application traffic.
  • the embodiments of the present application provide standardized measurement request messages, measurement response messages, path record information and delay record information, and a standardized network quality measurement protocol can be implemented based on these standardized message or information formats.
  • these standardized packets and information can carry not only the information of virtual network elements, but also the information of physical machines and tenants at the same time, so as to realize the network quality measurement of the three-layer network of physical network, overlay network and tenant network. Full coverage of end-to-end and intermediate node quality measurement, with the ability to quickly locate network problems.
  • the measurement system can be independently deployed outside the cloud network and communicate with network element devices in the cloud network, as shown in the figure 1d shown.
  • each subsystem in the measurement system can be distributed and deployed in a cloud network for implementation.
  • an embodiment of the present application further provides a cloud network with a network quality measurement function.
  • the cloud network 200 includes a physical network 201 and a virtual network 202 deployed between the physical networks 201 .
  • the virtual network 202 includes a multi-tenant network 2021.
  • the network of Tenant 1, Tenant 2 and Tenant 3 is taken as an example for illustration, but it is not limited thereto.
  • the physical network 201 includes physical machines 201a such as servers, cabinets, routers, and switches, and physical connecting lines 201b for realizing network connections between these physical machines, such as coaxial cables, network cables, and optical fibers.
  • the physical machine 201a includes physical server S1, physical server S2, physical server S3, and physical switches and physical routers connected between these physical servers as an example for illustration, but those skilled in the art should understand that the entire physical The network resources and network architecture included in the network 201 are not limited to this.
  • a virtualization technology is used to virtualize network resources in the physical network 201 , thereby obtaining a virtual network 202 borne on the physical network 201 .
  • the virtual network 202 includes a multi-tenant network 2021.
  • the networks of Tenant 1, Tenant 2, and Tenant 3 are used as examples for illustration, but not limited thereto. Further optionally, the networks of Tenant 1, Tenant 2 and Tenant 3 may be implemented as respective VPCs, but not limited thereto.
  • the network environments of different tenants are isolated from each other. Each tenant network includes end devices 202a visible to the tenant. These end devices 202a are carried on the physical machines 201a in the physical network 201. In FIG.
  • the end devices 202a are VMs.
  • these VMs are located on physical servers S1 and S2 in the physical network 201 .
  • the virtual network 202 further includes: a network element device 202b for performing traffic forwarding and network interconnection between different end devices 202a, and the network element device 202b may be a virtual switch or a virtual gateway.
  • the network element device 202b is taken as an example of a virtual switch or a virtual gateway for illustration, but it is not limited to this.
  • the network element device 202b belongs to the virtual network 202, but does not belong to any tenant network.
  • an Overlay network 2022 is implemented in the virtual network 202 by using the Overlay technology, and the Overlay network 2022 is between the physical network 201 and the multi-tenant network 2021, and the above-mentioned network element device 202b may be located at Implemented in Overlay Network 2022.
  • the overlay network 2022 is a bridge between the multi-tenant network 2021 and the physical network 201, so that the interconnection between the middle-end devices 202a of the multi-tenant network 2021 can escape the constraints of the physical network, and finally realize the flexible definition, on-demand allocation, and on-demand network resources. Conditions are provided for adjustment.
  • the entire cloud network 200 can be implemented as a three-layer network architecture including the physical network 201 , the overlay network 2022 and the multi-tenant network 2021 , but the division of the cloud network architecture is not limited to this.
  • the cloud network 200 in this embodiment further includes: a scheduling device 203 and an analysis device 204 .
  • the deployment positions of the scheduling device 203 and the analysis device 204 in the cloud network 200 are not limited.
  • the scheduling device 203 and the analyzing device 204 are deployed in the physical network 201 for implementation, for example, the two are deployed on the same physical machine 201a in the physical network 201 for implementation, or the two are deployed in the physical network 201 respectively.
  • One or more physical machines 201a can be added to the physical network 201 for deploying the scheduling device 203 and the analysis device 204; alternatively, the scheduling device 203 and the analysis device 204 It is directly deployed to the original physical machine 201 a in the physical network 201 .
  • the scheduling device 203 and the analyzing device 204 are deployed in other network locations in the virtual network 202 except the multi-tenant network 2021 for implementation, for example, both can be deployed in the overlay network 2022 for implementation.
  • the scheduling device 203 and the analysis device 204 are both deployed on the physical server S3 in the physical network 201 for illustration as an example, but it is not limited thereto.
  • each network element device 202b in the virtual network 202 can cooperate with the scheduling device 203 and the analyzing device 204 to measure the network quality in addition to the functions of traffic forwarding, packet processing, and network interconnection, which mainly refers to making a measurement request.
  • the processing operations of each network element device 202b will vary according to the location of the network element device 202b on the path to be measured. In this embodiment, the following will focus on the detailed description of the process in which the network element device 202b cooperates with the scheduling device 203 and the analysis device 204 to implement network quality measurement.
  • the scheduling device 203 perceives the measurement intent of the target tenant, generates a measurement rule adapted to the measurement intent, and delivers it to the source network element device, where the measurement rule includes the source device and the destination on the path to be measured
  • the source network element device is any network element device 202b on the path to be tested.
  • the source network element device is responsible for generating a measurement request message according to the measurement rule, and forwarding the measurement request message, where the measurement request message is used for the source network element device and the path to be measured to receive the measurement Other network element devices requesting packets generate measurement record information.
  • the source network element device can be any network element device on the path to be measured, which can be determined according to the measurement requirements.
  • the source network element device is the first network element device directly connected to the source end device on the path to be measured. In this case, the measurement request packet is forwarded to the target from the first network element device. network element equipment.
  • the embodiment of the present application does not limit the target network element device.
  • the target network element device is a second network element device that is directly connected to the destination device on the path to be tested.
  • the method for generating the measurement request message by the source network element device includes: generating a protocol header in the measurement request message according to the information of the source end device and the destination end device; The detailed information of the source network element device and the first timestamp are used to generate the measurement header in the measurement request packet; the measurement request packet is generated according to the protocol header and the measurement header; wherein, the first timestamp represents the source network element device The time when the measurement request message is sent to the outside world.
  • the source network element device or other network element device generates measurement record information in addition to generating the measurement request message and sending the measurement request message. Further optionally, the measurement record information includes route record information. Based on this, the source network element device or other network element device generates measurement record information, including: recording the identity of the tenant to which the network element itself belongs, detailed information of the network element itself, the second timestamp of the received measurement request packet, and the forwarding of the measurement request. at least one type of information in the third timestamp of the packet to generate path record information. The path record information is reported to the analysis device 204 by the source network element device or other network element devices.
  • the source network element device is further configured to: generate delay record information according to the first time stamp carried in the measurement reply message and the fourth time stamp of the received measurement reply message. The delay record information is reported to the analysis device 204 by the source network element device.
  • the analysis device 204 is configured to perform network quality analysis according to measurement record information generated by the source network element device and other network element devices.
  • the network quality analysis performed by the analysis device 204 according to the measurement record information reference may be made to the detailed implementation of the network quality analysis performed by the measurement and analysis subsystem 103 in the foregoing embodiment, which will not be repeated here.
  • scheduling equipment and analysis equipment are added, and the functions of network element equipment in the virtual network are expanded, so that the cloud network can sense the measurement intention of the tenant and perform network quality measurement based on the measurement intention.
  • it can free the tenant from the network quality measurement and simplify the network quality measurement work;
  • the bypass packet injection method there is no need to deploy the measurement service in the tenant's network environment, which is beneficial to reduce the impact on the tenant's network environment.
  • Intrusion in addition, the use of active packet injection can avoid dependence on the actual application traffic of the tenant, and the network quality can be measured even when the tenant does not generate actual application traffic.
  • the messages and information in the cloud network can simultaneously carry the information of virtual network elements, physical machines, and tenants, and can realize information on physical networks and overlay networks.
  • the network quality measurement of the three-layer network of the tenant network can support the full coverage of the network end-to-end and intermediate node quality, and has the ability to quickly locate network problems.
  • the cloud network of this embodiment supports multi-tenancy, the number of tenants is not limited, and it supports tenants to deploy multiple applications; in addition, this embodiment does not limit the link length of the cloud network, and can be flexibly deployed according to application requirements , adjustment, and the cloud network in this embodiment has the advantage of being easy to deploy, and can be flexibly and widely deployed according to application requirements.
  • an embodiment of the present application further provides a network quality measurement method, which is applicable to the measurement system or cloud network provided by the above-mentioned embodiments, but is not limited thereto. As shown in Figure 3a, the method includes:
  • the above-mentioned sensing the measurement intention of the target user in the cloud network includes: generating network topology change information of the target tenant according to the application requirement change information submitted by the target tenant; determining the target tenant's network topology change information according to the network topology change information. Measurement intent.
  • the above-mentioned generating a measurement rule adapted to the measurement intention includes: determining the source device and the destination device on the path to be measured according to the measurement intention of the target tenant and the network configuration information of the target tenant; The source device and destination device on the path to be tested generate measurement rules.
  • the above-mentioned determining the source device and the destination device on the path to be measured according to the measurement intention of the target tenant and the network configuration information of the target tenant includes: from the measurement intention of the target tenant, parsing out the target tenant.
  • the above-mentioned determining the destination device corresponding to the source device from the other end devices on the potential path includes: if the target tenant selects the full coverage measurement method, All devices are used as the destination device corresponding to the source device; if the target tenant chooses the selective measurement method, some potential paths are selected from all potential paths, and the other end device on some potential paths is used as the destination corresponding to the source device. end device.
  • the method of this embodiment further includes: displaying a human-computer interaction interface to the target tenant, and a measurement method option is set on the human-computer interaction interface for the target tenant to select; in response to the target tenant's selection of the measurement method option Operation, determine the measurement mode selected by the target tenant, and the measurement mode selected by the target tenant is the full coverage measurement mode or the selective measurement mode.
  • the above-mentioned generating the measurement request message according to the measurement rule includes: generating a protocol header in the measurement request message according to the information of the source device and the destination device; The detailed information of the network element device and the first timestamp are used to generate the measurement header in the measurement request packet; the measurement request packet is generated according to the protocol header and the measurement header; wherein, the first timestamp indicates that the measurement request packet is sent time.
  • the above-mentioned injecting the measurement request message into the path to be tested, so that at least some network element devices on the path to be tested generate measurement record information includes: sending the measurement request message to the path to be tested.
  • the source network element device on the device starts to forward the measurement request message to other network element devices on the path to be measured, so that the source network element device and other network element devices generate measurement record information.
  • a measurement execution subsystem can be bypassed next to each network element device in the cloud network, and the measurement execution subsystem generates a measurement request message and injects it into the network element device that has a bypass relationship with it;
  • injecting the measurement request message into the source network element device specifically refers to the process of sending the measurement request message to the source network element device after the measurement execution subsystem that has a bypass relationship with the source network element device generates the measurement message.
  • each network element device in the cloud network can be expanded, so that each network element device has the function of generating a measurement request message according to the measurement rule; based on this, a measurement request is injected into the source network element device.
  • the message specifically refers to the process in which the source network element device generates a measurement request message based on the extended message generation function.
  • the source network element device or other network element device generates measurement record information, including: recording the identity of the tenant to which the network element itself belongs, detailed information of the network element itself, and the first number of the measurement request message received. Two timestamps, at least one type of information in the third timestamp of the measurement request packet is forwarded, so as to generate path record information.
  • generating the measurement record information by the source network element device further includes: receiving a measurement reply message sent by other network element devices, where the measurement reply message includes the first timestamp; and replying according to the measurement The first time stamp in the message and the fourth time stamp of the received measurement reply message are used to generate delay record information.
  • performing the network quality analysis according to the measurement record information generated by at least part of the network element equipment includes: according to the above-mentioned delay record information and path record information, in the path to be measured, the target tenant and the cloud network. Analyze network latency and packet loss in at least one dimension.
  • an embodiment of the present application also provides a method for generating measurement rules.
  • the method is described from the perspective of a scheduling subsystem or a scheduling device. As shown in Fig. 3b, the method includes :
  • the measurement rule includes the source end device and the destination end device on the path to be measured;
  • the source network element device may be any network element device on the path to be tested, for example, it may be the first network element device directly connected to the source end device on the path to be tested.
  • an embodiment of the present application also provides a message transmission method.
  • the method is mainly described from the perspective of a network element device in a cloud network. As shown in FIG. 3c, the method includes:
  • the source network element device receives the measurement rule, the measurement rule includes the source end device and the destination end device on the path to be tested, and the source network element device is the network element device on the path to be tested;
  • the source and generation method of the measurement rule are not limited.
  • the measurement rule may be generated in the manner in the foregoing embodiment.
  • the execution subject of each step of the method provided in the above-mentioned embodiments may be the same device, or the method may also be executed by different devices.
  • the execution subject of steps 31a to 33a may be device A; for another example, the execution subject of step 31a may be device A, and the execution subject of step 32a may be device B; and so on.
  • FIG. 4 is a schematic structural diagram of an apparatus for measuring network quality according to an exemplary embodiment of the present application. As shown in FIG. 4 , the apparatus includes: an intention perception module 41 , a rule generation module 42 , a message generation module 43 , a message injection module 44 and a quality analysis module 45 .
  • the intention sensing module 41 is used for sensing the measurement intention of the target tenant in the cloud network.
  • the rule generation module 42 is configured to generate a measurement rule adapted to the measurement intention sensed by the intention sensing module 41 , and the measurement rule includes the source end device and the destination end device on the path to be measured.
  • the message generation module 43 is configured to generate a measurement request message according to the measurement rules generated by the rule generation module 42 .
  • the message injection module 44 is configured to inject the measurement request message generated by the message generation module 43 into the path to be tested, so that at least some network element devices on the path to be tested generate measurement record information.
  • the quality analysis module 45 is configured to perform network quality analysis according to the measurement record information generated by at least part of the network element devices.
  • the intention sensing module 41 is specifically configured to: generate network topology change information of the target tenant according to the application requirement change information submitted by the target tenant; determine the target tenant according to the network topology change information measurement intent.
  • the rule generation module 42 is specifically configured to: determine the source device and the destination device on the path to be tested according to the measurement intention of the target tenant and the network configuration information of the target tenant; The source device and the destination device generate measurement rules.
  • the rule generation module 42 is specifically configured to: parse out the source end device to be measured from the measurement intention of the target tenant; The network topology where it is located, combined with the network configuration information of the target tenant, obtains a potential path that has an access relationship with the source device; from the other end devices on the potential path, determines the destination device corresponding to the source device.
  • the rule generation module 42 determines the destination device corresponding to the source device, it is specifically used for: if the target tenant selects the full coverage measurement method, the other end devices on all potential paths are regarded as corresponding to the source device. If the target tenant chooses the selective measurement method, select some potential paths from all potential paths, and take the other end device on the part of the potential paths as the destination end device corresponding to the source end device.
  • the apparatus further includes: a configuration module 46 for displaying a human-computer interaction interface to the target tenant, and the human-computer interaction interface is provided with measurement mode options for all The target tenant is selected; in response to the selection operation initiated by the target tenant for the measurement mode option, the measurement mode selected by the target tenant is determined, and the measurement mode selected by the target tenant is a full coverage measurement mode or a selective measurement mode.
  • the message injection module 44 is specifically configured to: send the measurement request message to the source network element device on the path to be tested, and start from the source network element device to other network element devices on the path to be tested. Forward the measurement request message, so that the source network element device and other network element devices generate measurement record information.
  • the message generation module 43 is specifically configured to: generate the protocol header in the measurement request message according to the information of the source device and the destination device; according to the identifier of the target tenant; , the detailed information of the source network element device and the first timestamp, to generate the measurement header in the measurement request message; according to the protocol header and the measurement header, generate the measurement request message, the first The timestamp indicates the time when the measurement request message is sent.
  • the network quality measurement device of this embodiment can be distributed and deployed in the cloud network.
  • the intent sensing module 41, the rule generation module 42, and the quality analysis module 45 can be independently deployed in the cloud network, while the packet generation module 41 can be independently deployed in the cloud network.
  • the module 43 and the message injection module 44 may be implemented in network element devices deployed in the cloud network. Further optionally, in the case where the message generation module 43 and the message injection module 44 are implemented in a network element device, the message generation module 43 and the message injection module 44 may be implemented as one module.
  • the message generation module 43 or the message injection module 44 is further configured to generate measurement record information. Specifically, the message generation module 43 or the message injection module 44 records the identity of the tenant to which the network element itself belongs, the detailed information of the network element itself, the second timestamp when the measurement request message is received, and the time stamp for forwarding the measurement request message. at least one type of information in the third time stamp to generate path record information.
  • the message generation module 43 or the message injection module 44 may also receive a measurement reply message sent by other network element devices, where the measurement reply message includes the first timestamp; The first timestamp of the received measurement reply message and the fourth timestamp of the received measurement reply message are used to generate delay record information.
  • the quality analysis module 45 is specifically configured to: analyze the network delay and packet loss in at least one dimension of the path to be tested, the target tenant, and the cloud network according to the above-mentioned delay record information and path record information. Rate.
  • the memory 51 is used to store computer programs and may be configured to store various other data to support operations on the cloud computing device. Examples of such data include instructions, messages, pictures, videos, etc. for any application or method operating on the cloud computing device.
  • the processor 52 coupled with the memory 51, is used for executing the computer program in the memory 51, so as to: perceive the measurement intention of the target user in the cloud network, and generate a measurement rule adapted to the measurement intention, the measurement rule Including the source end device and the destination end device on the path to be tested; generate a measurement request message according to the measurement rule, and inject the measurement request message into the path to be tested, so that at least part of the network element devices on the path to be tested generating measurement record information; and performing network quality analysis according to the measurement record information generated by at least part of the network element equipment.
  • the processor 52 when sensing the measurement intention, is specifically configured to: generate network topology change information of the target tenant according to the application requirement change information submitted by the target tenant; according to the network topology change information, Determine the measurement intent of the target tenant.
  • the processor 52 when generating the measurement rule, is specifically configured to: determine the source end device and the destination end device on the path to be measured according to the measurement intention of the target tenant and the network configuration information of the target tenant; The source device and destination device on the path to be tested generate measurement rules.
  • the processor 52 is specifically configured to: parse out the source end device to be measured from the measurement intent of the target tenant; Based on the network topology of the target tenant, obtain the potential path that has an access relationship with the source device in combination with the network configuration information of the target tenant; determine the destination device corresponding to the source device from the other devices on the potential path.
  • the processor 52 determines the destination device corresponding to the source device, the processor 52 is specifically configured to: if the target tenant selects the full coverage measurement method, the other end devices on all potential paths are regarded as the destination device corresponding to the source device. Destination device; if the target tenant selects the selective measurement method, select some potential paths from all potential paths, and use the other end device on the part of the potential paths as the destination end device corresponding to the source end device.
  • the processor 52 is further configured to: display a human-computer interaction interface to the target tenant, and the human-computer interaction interface is provided with measurement mode options for the target tenant to choose; The selection operation initiated by the measurement mode option determines the measurement mode selected by the target tenant, and the measurement mode selected by the target tenant is a full coverage measurement mode or a selective measurement mode.
  • the processor 52 when injecting the measurement request message into the path to be tested, is specifically configured to: send the measurement request message to the source network element device on the path to be tested, from the source network element device. Start to forward the measurement request message to other network element devices on the path to be measured, so that the source network element device and other network element devices generate measurement record information.
  • the processor 52 when generating the measurement request message, is specifically configured to: generate a protocol header in the measurement request message according to the information of the source end device and the destination end device; the identifier of the target tenant, the detailed information of the source network element device and the first timestamp, to generate the measurement header in the measurement request message; and according to the protocol header and the measurement header, to generate the measurement request message,
  • the first time stamp indicates the time when the measurement request message is sent.
  • the cloud computing device in this embodiment may be deployed and implemented in a cloud network, and specifically, the functions of generating and injecting test packets may be implemented in a network element device in the cloud network.
  • the processor 52 is also configured to generate measurement record information.
  • the processor 52 is specifically configured to: record the identity of the tenant to which the network element where it is located, the detailed information of the network element itself, the second time stamp of receiving the measurement request message, and the third time stamp of the forwarding measurement request message. at least one kind of information to generate path record information.
  • the processor 52 is further configured to: receive a measurement reply message sent by other network element devices, where the measurement reply message includes the first timestamp; The first timestamp of the received measurement reply message and the fourth timestamp of the received measurement reply message are used to generate delay record information.
  • the processor 52 when the processor 52 performs network quality analysis, it is specifically configured to: analyze the path to be tested, the target tenant, and the cloud network in at least one dimension according to the above-mentioned delay record information and path record information. Network latency and packet loss rate.
  • the cloud computing device further includes: a power supply component 54 and other components. Only some components are schematically shown in FIG. 5 , which does not mean that the cloud computing device only includes the components shown in FIG. 5 .
  • the embodiments of the present application further provide a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the processor can implement the steps in the foregoing method embodiments.
  • an embodiment of the present application further provides a computer program product, including a computer program/instruction, when the computer program/instruction is executed by a processor, the processor can implement each step in the above method embodiment.
  • the memory in the above-described embodiments may be implemented by any type of volatile or non-volatile memory device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.
  • SRAM Static Random Access Memory
  • EEPROM Electrically Erasable Programmable Read-Only Memory
  • EPROM Erasable Programmable Read Only Memory
  • PROM Programmable Read Only Memory
  • ROM Read Only Memory
  • Magnetic Memory Flash Memory
  • Magnetic or Optical Disk any type of volatile or non-volatile memory device or a combination thereof
  • the communication components in the above embodiments are configured to facilitate wired or wireless communication between the device where the communication components are located and other devices.
  • the device where the communication component is located can access a wireless network based on a communication standard, such as WiFi, a mobile communication network such as 2G, 3G, 4G/LTE, 5G, or a combination thereof.
  • the communication component receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communication assembly further includes a near field communication (NFC) module to facilitate short-range communication.
  • the NFC module may be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • a power supply assembly may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the equipment in which the power supply assembly is located.
  • the embodiments of the present application may be provided as a method, a system, or a computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
  • computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions
  • the apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.
  • a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
  • processors CPUs
  • input/output interfaces network interfaces
  • memory volatile and non-volatile memory
  • Memory may include forms of non-persistent memory, random access memory (RAM) and/or non-volatile memory in computer readable media, such as read only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
  • RAM random access memory
  • ROM read only memory
  • flash RAM flash memory
  • Computer-readable media includes both persistent and non-permanent, removable and non-removable media, and storage of information may be implemented by any method or technology.
  • Information may be computer readable instructions, data structures, modules of programs, or other data.
  • Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), Flash Memory or other memory technology, Compact Disc Read Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, Magnetic tape cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device.
  • computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

Landscapes

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

Abstract

本申请实施例提供一种云网络、用于云网络的测量系统、方法、设备及存储介质。在本申请实施例中，针对云网络，自动感知云网络中租户的测量意图，根据测量意图生成测量规则，基于该测量规则采用旁路注包的方式向待测路径上的网元设备内部注入测量请求报文，借助于测量请求报文在经过不同网元设备时产生的测量记录信息进行网络质量分析，可简单、有效、准确地测量出云网络的网络质量。进一步，通过对租户测量请求的意图化，可极大地简化测量过程；通过旁路注包的方式，有利于降低了对租户网络环境的侵入；同时，采用主动注包方式，可避免对租户实际应用流量的依赖，即使在租户没有产生实际应用流量的情况下也可以对网络质量进行测量。

Description

云网络、用于云网络的测量系统、方法、设备及存储介质
本申请要求2021年05月07日递交的申请号为202110496812.X、发明名称为“云网络、用于云网络的测量系统、方法、设备及存储介质”中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及云计算技术领域，尤其涉及一种云网络、用于云网络的测量系统、方法、设备及存储介质。
背景技术
云网络(Cloud Network)是一个融合了物理网络和虚拟网络，虚拟网络中包括多租户网络的复杂网络，它承载了租户应用以及应用之间的连接和通信，租户的应用请求会经过多层网络的处理和传输之后到达目的端。
云网络的网络质量对租户应用有直接的影响。然而，由于云网络的租户、网络拓扑具有灵活多变的特点，现有针对物理网络的质量探测技术并不适用于云网络。在这个情况下，如何简单准确地测量出云网络的网络质量非常关键。
发明内容
本申请的多个方面提供一种云网络、用于云网络的测量系统、方法、设备及存储介质，用以简单、准确地测量云网络的网络质量。
本申请实施例提供一种云网络，包括：物理网络和承载于所述物理网络之上的虚拟网络；所述虚拟网络包括多租户网络，以及负责在所述多租户网络中不同端设备之间进行流量转发和互联的网元设备，所述云网络还包括：调度设备和分析设备；所述调度设备，用于感知目标租户的测量意图，生成与所述测量意图适配的测量规则并下发给源网元设备，所述测量规则包括待测路径上的源端设备和目的端设备，所述源网元设备是所述待测路径上的网元设备；所述源网元设备，用于根据所述测量规则生成测量请求报文，并将所述测量请求报文转发出去，所述测量请求报文用于供所述源网元设备和所述待测路径上收到所述测量请求报文的其它网元设备生成测量记录信息；所述分析设备，用于根据所述源网元设备和其它网元设备生成的测量记录信息进行网络质量分析。
本申请实施例还提供一种用于云网络的测量系统，包括：调度子系统、至少一个测量执行子系统和测量分析子系统；所述调度子系统，用于感知云网络中目标租户的测量意图，生成与所述测量意图适配的测量规则并下发给目标测量执行子系统，所述测量规则包括待测路径上的源端设备和目的端设备；所述目标测量执行子系统与所述待测路径对应；所述目标测量执行子系统，用于根据所述测量规则生成测量请求报文，并将所述测量请求报文注入到所述待测路径上，以使所述待测路径上的至少部分网元设备对所述测量请求报文进行转发并产生测量记录信息；所述测量分析子系统，用于根据所述至少部分网元设备产生的测量记录信息进行网络质量分析。
本申请实施例还提供一种网络质量测量方法，包括：感知云网络中目标用户的测量意图，并生成与所述测量意图适配的测量规则，所述测量规则包括待测路径上的源端设备和目的端设备；根据所述测量规则生成测量请求报文，并将所述测量请求报文注入到所述待测路径上，以使所述待测路径上的至少部分网元设备生成测量记录信息；根据所述至少部分网元设备生成的测量记录信息进行网络质量分析。
本申请实施例还提供一种云计算设备，包括：存储器和处理器；所述存储器，用于存储计算机程序；所述处理器，用于执行所述计算机程序，以用于实现本申请方法实施例中的步骤。
本申请实施例还提供一种存储有计算机程序的计算机可读存储介质，当所述计算机程序被处理器执行时，致使所述处理器能够实现本申请方法实施例中的步骤。
本申请实施例还提供一种计算机程序产品，包括计算机程序/指令，当所述计算机程序/指令被处理器执行时，致使所述处理器能够实现本申请方法实施例中的步骤。
在本申请实施例中，针对云网络，通过感知云网络中租户的测量意图，根据测量意图生成测量规则，基于该测量规则采用旁路注包的方式向待测路径上的网元设备内部注入测量请求报文，借助于测量请求报文在经过不同网元设备时产生的测量记录信息进行网络质量分析，可简单、有效、准确地测量出云网络的网络质量。
进一步，通过对租户测量请求的意图化，可极大地简化测量过程，而且租户对测量过程无感知，无需对复杂测量规则进行管理操作，有利于提高租户体验度；另外，通过旁路注包的方式，无需在租户的网络环境中部署测量服务，有利于降低了对租户网络环境的侵入；同时，由于本申请实施例是主动注包，可避免对租户实际应用流量的依赖，即使在租户没有产生实际应用流量的情况下也可以对网络质量进行测量。
附图说明
此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1a为本申请示例性实施例提供的一种用于云网络的测量系统的结构示意图；
图1b为本申请示例性实施例提供的向待测路径上注入测量请求报文生成测量记录信息的过程示意图；
图1c为本申请示例性实施例提供的测量分析子系统的内部结构示意图；
图1d为本申请示例性实施例提供的测量分析子系统独立于云网络部署时的使用状态示意图；
图2a为本申请示例性实施例提供的一种云网络的结构示意图；
图2b为本申请示例性实施例提供的另一种云网络的结构示意图；
图3a为本申请示例性实施例提供的一种网络质量测量方法的流程示意图；
图3b为本申请示例性实施例提供的一种测量规则生成方法的流程示意图；
图3c为本申请示例性实施例提供的一种报文传输方法的流程示意图；
图4为本申请示例性实施例提供的一种网络质量测量装置的结构示意图；
图5为本申请示例性实施例提供的一种云计算设备的结构示意图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚，下面将结合本申请具体实施例及相应的附图对本申请技术方案进行清楚、完整地描述。显然，所描述的实施例仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
针对现有无法对云网络的网络质量进行有效测量的技术问题，在本申请一些实施例中，提供一种测量系统，该测量系统可用于对云网络的网络质量进行测量，但并不限于此。该测量系统也可以应用于物理网络，用于针对物理网络进行网络质量的测量。在将该测量系统应用于云网络的情况下，该测量系统可感知云网络中租户的测量意图，根据测量意图生成测量规则，基于该测量规则采用旁路注包的方式向待测路径上的网元设备内部注入测量请求报文，进而借助测量请求报文在经过不同网元设备时产生的测量记录信息进行网络质量分析，可简单、有效、准确地测量出云网络的网络质量。在本申请下 面实施例中，重点以将测量系统应用于云网络为例进行说明。
首先，对云网络进行简单介绍：云网络是一种融合了物理网络和虚拟网络，虚拟网络又包括多租户网络的复杂网络，即云网络包括物理网络和虚拟网络。其中，物理网络包括服务器、机柜、路由器、交换机等物理机，以及用于实现这些物理机之间的网络连接的物理连接线，例如同轴线缆、网线、光纤等。虚拟网络承载于物理网络之上，是在物理网络的基础上基于虚拟化技术实现的一种逻辑网络，可将物理网络资源进行虚拟化，使得物理网络资源升级为虚拟化、可动态分配的虚拟化资源。
在本申请实施例中，虚拟网络包括多租户网络，不同租户的网络环境之间相互隔离，每个租户网络包括但不限于：在该云网络中可识别为该租户的一切数据，比如在云网络中为该租户创建的账户与统计信息(Accounting Data)，以及在云网络中为该租户设置的各式数据和租户所配置的虚拟化的端设备等各种虚拟化资源。在本申请实施例中，租户网络中的端设备是指可以发起或终结网络流量的虚拟化设备，通常情况下端设备可以为租户提供完全隔离的网络环境，负责承载租户的应用，这些应用会进行网络流量(如各种报文)的收发，在实现形态上可以是但不限于：虚拟机(VM)、容器、弹性实例(ECS)、用于后续网元设备(如虚拟交换机)的配置对象、云数据库等。其中，租户(Tenant)泛指使用云网络或云网络中各种资源的用户。在本实施例中，并不限定租户网络的实现形态，例如可以实现为虚拟专有云(Virtual Private Cloud，VPC)，对租户来说，可以在其VPC中部署自己的VM、云数据库等端设备。
除了多租户网络之外，虚拟网络还包括在多租户网络中不同端设备之间实现流量转发和网络互联的网元设备。在本申请实施例中，网元设备主要用于在不同端设备之间进行流量转发，进行报文处理，实现不同端设备之间的网络互联，例如可以是虚拟交换机、虚拟网关等。在一可选实施例中，可以采用叠加(Overlay)技术在虚拟网络中实现Overlay网络，Overlay网络介于物理网络和租户网络之间，上述网元设备可位于Overlay网络中实现。Overlay网络是租户网络与物理网络之间的桥梁，使得租户网络中端设备之间的互联跳出物理网络的束缚，为最终实现网络资源的灵活定义、按需分配、按需调整提供了条件。这样整个云网络可实现为包括物理网络、Overlay网络和多租户网络的三层网络架构，但对云网络架构的划分并不限于此。无论是哪种网络架构，通过将虚拟网络中各种资源的配置信息进行包装最终以配置对象的方式呈现给租户，租户可以根据自身需求对配置对象进行相应配置，从而得到符合自己需求的端设备、子网络等资源。
由上述对云网络的介绍可知，对云网络进行网络质量探测不再局限在物理机或物理 机之间，而是需要深入到虚拟化的端设备(如虚拟机)内部，而虚拟化的端设备是逻辑节点，其与物理机的部署、拓扑都有较大差异。在云网络中，相比于物理网络的拓扑结构，租户网络的拓扑结构是灵活、多变的，根据租户需求的不同，这些虚拟化的端设备(如虚拟机)可灵活地增加、减少，端设备之间以及端设备与其它网络(例如专有网络)之间的互联关系也可以根据租户的需求灵活改变。如果由租户直接对云网络的网络质量进行探测，租户需要进行复杂的测量工作，包括部署测量服务，管理测量对象以及测量规则等，而且这些工作还需要根据虚拟网络的拓扑变化不断适应性地变化，这对租户来说是较为繁重的工作，网络质量测量难度较大。而本申请实施例提供的测量系统可解决云网络的网络质量测量难的技术问题。
接下来，对本申请实施例提供的测量系统进行介绍：如图1a所示，本申请实施例提供的可用于云网络的测量系统100包括：调度子系统101、至少一个测量执行子系统102和测量分析子系统103。其中，每个测量执行子系统102分别与调度子系统101和测量分析子系统103通信连接。本实施例中，调度子系统101、测量执行子系统102和测量分析子系统103可以采用软件形态实现，在使用测量系统100之前，可以部署调度子系统101、测量执行子系统102和测量分析子系统103各自对应的软件代码。在本实施例中，并不限定调度子系统101、测量执行子系统102和测量分析子系统103的部署位置，可根据应用需求部署在合适的网络位置。
在测量系统应用于云网络的情况下，如图1a所示，调度子系统101一方面用于感知云网络中目标租户的测量意图(measurement intent)，生成与测量意图适配的测量规则；另一方面，通过对测量执行子系统102进行调度，将测量规则下发给合适的测量执行子系统102，以使测量执行子系统102根据测量规则执行与网络质量测量相关的操作并产生测量记录信息；进而，由测量分析子系统103根据测量记录信息进行网络质量分析，产生网络质量分析结果。进一步可选地，如图1a所示，网络质量分析结果还可以反馈给调度子系统101，进一步影响测量执行子系统102对目标租户测量意图的感知过程。其中，调度子系统101负责感知目标租户的测量意图并根据测量意图生成与之适配的测量规则，可将目标租户从网络质量测量中解放出来，目标租户无需维护和管理测量规则，有利于提高目标租户对云网络的使用体验。
其中，目标租户是指云网络中的租户，其数量可以是一个或多个，可以是云网络中的全部租户，也可以是云网络中的部分租户，具体可以是网络质量测量需求而定。例如，如果是从云网络提供商的角度出发，需要了解整个云网络的网络质量，则可以将云网络 中的全部租户作为目标租户，感知每个租户的测量意图，针对每个租户进行网络质量测量，从而得到整个云网络的网络质量。又例如，如果是从租户的角度出发，只需了解这些租户(这些租户可以是提出网络质量测量要求的租户，或者是VIP租户，或者是网络质量低于设定标准的租户等)自身网络的网络质量，则可以将这些租户作为目标用户，感知这些租户的测量意图，针对这些租户进行网络质量测量，从而满足这些租户的网络质量测量需求。又例如，如果是从区域的角度出发，需要了解云网络覆盖到的某个区域的网络质量，则可以将该区域内的租户作为目标租户，感知这些租户的测量意图，针对这些租户进行网络质量测量，从而得到该区域内的网络质量。又例如，如果从应用类别的角度出发，需要了解某一类应用涉及的网络质量，则可以将部署这类应用的租户作为目标租户，感知这些租户的测量意图，针对这些租户进行网络质量测量，从而得到该类应用对应的网络质量。
在本实施例中，调度子系统101感知目标租户的测量意图；目标租户的测量意图是对目标租户的测量请求的意图化，能够反映出目标租户需要什么样的测量请求，该测量意图可以采用更加切近用户的语言表达或描述方式。可以理解为：该测量意图是以接近用户语言表达或者描述方式对目标租户的测量请求进行的表达。这里的测量请求是指请求测量云网络的网络质量的请求。
在本实施例中，调度子系统101可以采用各种方式来感知目标租户的测量意图，对此不做限定。在一可选实施例中，调度子系统101可以监测目标租户的网络拓扑是否发生变化；在目标租户的网络拓扑未发生变化的情况下，目标租户的测量意图一般不会变化；在目标租户的网络拓扑发生变化的情况下，例如目标租户新购买了一台虚拟机，删除了已有虚拟机或者更改了虚拟机的配置信息等，对网络质量的测量意图通常会发生变化；基于此，调度子系统101可以根据目标租户的网络拓扑变化信息来感知目标租户的测量意图。具体地，调度子系统101可以获取目标租户提交的应用需求变化信息，根据目标租户提交的应用需求变化信息，生成目标租户的网络拓扑变化信息；根据目标租户的网络拓扑变化信息，确定目标租户的测量意图。其中，目标租户的应用需求变化，会直接影响目标租户的网络拓扑。例如，因为下游用户量的增长，目标租户需要增加一台虚拟机，则目标租户的网络拓扑会因为新增一台虚拟机而发生变化，该台虚拟机可能需要与负载均衡器互联，或者该台虚拟机需要接入专线，或者该台虚拟机需要与已有其它虚拟机互联，这会导致目标租户对网络质量的测量意图发生变化。在本实施例中，并不限定调度子系统101获取目标租户提交的应用需求变化信息的实施方式，下面举例说明：
在一可选实施例A1中，调度子系统101可以实现为目标租户与云网络提供商进行交互的门户，目标租户一开始提交的应用需求信息或后续提交的应用需求变化信息都可以通过调度子系统101进行提交，以便于云网络提供商为目标租户提供相应的云计算服务。基于此，调度子系统101可以直接接收目标租户提交的应用需求变化信息。在本实施例中，调度子系统101可以面向目标租户提供web网页、应用页面或命令窗，目标租户在自己使用的终端设备上打开调度子系统101提供的web网页、应用页面或命令窗，通过web网页、应用页面或命令窗向调度子系统101提交应用需求变化信息。
在另一可选实施例A2中，云网络的虚拟网络中还包括网络控制器，该网络控制器是租户与云网络提供商之间进行交互的门户，在实现方式上，网络控制器可以是一些分布式部署的软件，例如可以是API。基于网络控制器，目标租户可以向云网络提供商提交应用需求信息或应用需求变化信息。在本实施例中，网络控制器可以面向目标租户提供web网页、应用页面或命令窗，目标租户在自己使用的终端设备上打开网络控制器提供的web网页、应用页面或命令窗，通过web网页、应用页面或命令窗向网络控制器提交应用需求变化信息。基于此，调度子系统101与云网络中的网络控制器进行通信连接，从网络控制器中获取目标租户提交的应用需求变化信息。
进一步可选地，在实施例A2中，调度子系统101可以旁挂在网络控制器的旁边，通过旁路采集方式将目标租户提交的应用需求变化信息从网络控制器中分流至调度子系统101中。或者，在实施例A2中，网络控制器可以面向调度子系统101提供订阅服务，调度子系统101可以注册为网络控制器的订阅者，用于向网络控制器订阅租户提交的应用需求变化信息，这样，当网络控制器接收到任何租户提交的应用需求变化信息，可以主动将租户提交的应用需求变化信息提供给调度子系统101。
无论采用上述哪种实施例提供的方式，在获取目标租户提交的应用需求变化信息之后，调度子系统101可以根据该应用需求变化信息，结合所维护的云网络的网络拓扑，生成目标租户的网络拓扑变化信息；进而根据该网络拓扑变化信息，确定目标租户的测量意图。例如，在目标租户增加一台虚拟机的情况下，调度子系统101可以在目标租户原有网络拓扑的基础上增加新的虚拟机以及与新虚拟机相关的网络连接，得到目标租户的网络拓扑变化信息；进而，感知目标租户有可能需要对已有虚拟机与该新增虚拟机之间的网络质量进行测量，则测量意图可表示为：测量已有虚拟机Vm1与新增虚拟机Vm2之间的网络质量。又例如，在目标租户根据应用需求为已有虚拟机新配置公网IP地址的情况下，调度子系统101可以在目标租户原有网络拓扑的基础上增加已有虚拟机与公网 之间的网络连接，得到目标租户的网络拓扑变化信息；进而，感知目标租户的虚拟机需要访问该公网IP对应的公网，有必要对虚拟机到公网的网络质量进行探测，假设该公网为公网-12345，则测量意图可表示为：测量到公网-12345的网络质量(measure network-12345)。
在感知到目标租户的测量意图之后，调度子系统101可生成与测量意图适配的测量规则，该过程是指对测量意图进行解析，进而得到进行网络质量测量所需的相关信息，并按照测量执行子系统102可识别的信息格式对这些信息进行组织的过程，即最终得到的测量规则。其中，测量规则包括根据网络质量测量需求对云网络进行网络质量测量所需的相关信息，这些信息至少包括能够表示需要对哪些端设备之间的路径进行网络质量的测量的信息，例如可以包括但不限于：源端设备和目的端设备的标识信息，源端设备和目的端设备可以唯一确定一条网络路径，在本申请实施例中，将由源端设备与目的端设备限定的网络路径称为待测量路径。这里的标识信息可以是任何能够唯一标识源端设备或目的端设备的信息，例如可以是IP地址、MAC地址或设备序列号等。其中，源端设备和目的端设备可以是一组或多组，具体视网络质量测量需求而定。进一步可选地，测量规则中还可以包括源端设备和目的端设备之间进行网络通信所采用的通信协议，例如UDP协议、TCP协议等，以便于采用源端设备和目的端设备支持的通信协议对两者之间的路径进行网络质量测量。以测量意图是测量已有虚拟机Vm1与新增虚拟机Vm2之间的网络质量为例，则根据该测量意图生成的测量规则可表示为：源端设备(Src)：Vm1，目的端设备(Dst)：Vm2，通信协议(protocol)：UDP。
在本申请实施例中，并不限定调度子系统101生成测量规则的实施方式。在一可选实施例中，调度子系统101可以根据目标租户的测量意图，结合目标租户的网络配置信息，确定待测路径上的源端设备和目的端设备；进而，根据待测路径上的源端设备和目的端设备，生成测量规则。其中，目标租户的网络配置信息是指目标租户通过租户网络层进行配置的各种信息，例如包括但不限于：路由表项信息、访问控制列表(ACL)信息、网元设备的信息、网络拓扑信息等。进一步，在确定源端设备和目的端设备时，可以从目标租户的测量意图中，解析出待测量的源端设备，例如可以是目标租户新增的虚拟机，或者需要访问的公网或专线等；之后，从整个云网络的拓扑中确定源端设备所在的网络拓扑，根据源端设备所在的网络拓扑，结合目标租户的网络配置信息，获取与源端设备存在访问关系的潜在路径；其中，潜在路径可以是一条或多条；潜在路径一端的端设备为源端设备，并从潜在路径上的另一端设备中，确定与该源端设备对应的目的端 设备。在本实施例中，并不限定从潜在路径上的另一端设备中，确定与源端设备对应的目的端设备的实施方式，下面举例说明：
在一可选实施例B1中，可以默认将全部潜在路径上的另一端设备均作为与源端设备对应的目的端设备，也就意味着，所有潜在路径均会被视为待测路径，通过测量至少可以了解到所有潜在路径的网络质量。
在另一可选实施例B2中，可以设定每次测量所允许的最大路径数量，记为N，N是固定值，例如可以是1、2、3或4等整整数值，则在潜在路径为多条的情况下，可以随机从中选择数量不大于N的至少一条潜在路径，将被选中的至少一条潜在路径上的另一端设备作为与源端设备对应的目的端设备，相应地，被选中的至少一条潜在路径会被视为待测路径，通过测量至少可以了解到被选中的至少一条潜在路径的网络质量。
在又一可选实施例B3中，允许目标租户选择所使用的测量方式，用户可选择的测量方式包括但不限于：全覆盖测量方式和选择性测量方式。其中，全覆盖测量方式是指对所有存在的潜在路径均进行测量的方式，选择性测量方式是指有选择性地对部分潜在路径进行测量的方式。在该可选实施例中，目标用户可以选择所使用的测量方式；若目标租户选择全覆盖测量方式，在可以将全部潜在路径上的另一端设备均作为与源端设备对应的目的端设备，即所有潜在路径均会被视为待测路径，通过测量至少可以了解到所有潜在路径的网络质量；若目标租户选择选择性测量方式，则可以从全部潜在路径中选择部分潜在路径，将所选择的部分潜在路径上的另一端设备作为与源端设备对应的目的端设备，即被选择的部分潜在路径会被视为待测路径，通过测量至少可以了解到被选择的部分潜在路径的网络质量。
关于如何从全部潜在路径中选择部分潜在路径，本申请实施例对此不做限定，例如可以随机选择部分潜在路径，或者，根据每次测量所允许的最大路径数量N选择部分潜在路径，或者，也可以允许目标租户设定选择条件，例如可以是另一端设备的类型、区域位置、入网时间等，根据目标租户设定的选择条件从中选择部分潜在路径。
进一步可选地，在实施例B3中，调度子系统101可以向目标租户展示人机交互界面，在该人机交互界面上设置有测量方式选择项，该选择项至少包括全覆盖测量方式和选择性测量方式，以供目标租户进行选择；响应于目标租户对该测量方式选项发起的选择操作，确定目标租户选择使用的测量方式，其中，目标租户选择使用的测量方式或者是全覆盖测量方式，或者是选择性测量方式。除了该实施方式之外，调度子系统101也可以通过云网络中的网络控制器，获取目标租户选择使用的测量方式。具体地，网络控 制器可以向目标租户展示人机交互界面，在该人机交互界面上设置有测量方式选择项，该选择项至少包括全覆盖测量方式和选择性测量方式，以供目标租户进行选择；响应于目标租户对该测量方式选项发起的选择操作，确定目标租户选择使用的测量方式，并提供给调度子系统101。在本实施例中，并不对人机交互界面的实现形式进行限定，例如可以是web网页、应用页面或命令窗等。
在得到与测量意图适配的测量规则之后，调度子系统101可以将测量规则下发给测量执行子系统102，由测量执行子系统102根据测量规则执行与网络质量测量相关的操作。在本实施例中，测量执行子系统102可以是一个，也可以是多个。无论测量执行子系统102是一个还是多个，本实施例均不限定测量执行子系统102的部署位置，测量执行子系统102可以部署在一条路径上除端设备之外的任何位置，即在本实施例中，不在租户的端设备中部署测量执行子系统102，这样可以减轻网络质量测量对租户网络环境的侵入，使得租户无需对复杂的测量规则进行管理和维护。
在测量执行子系统102为一个的情况下，无论待测路径是哪条或哪些条，调度子系统101会直接将测量规则下发给该测量执行子系统102，均由该测量执行子系统102针对待测路径执行与网络质量测量相关的操作。在测量执行子系统102为多个情况下，本实施例的调度子系统101还负责对测量执行子系统102进行调度，以便于采用与待测路径适配的测量执行子系统102对其进行网络质量测量。进一步，在测量执行子系统102为多个的情况下，可以采用分布式部署方式对多个测量执行子系统102进行部署，即多个测量执行子系统102会部署在不同的位置，另外，不同测量执行子系统102所占用的物理资源也可以有所不同。鉴于此，调度子系统101在调度测量执行子系统102时，可以优先选择位置上与待测路径更加靠近的测量执行子系统，和/或，选择执行性能更优的测量执行子系统102，和/或，选择负载量更少的测量执行子系统102，等等，对此不做限定。
具体地，调度子系统101可以将上述测量规则下发给目标测量执行子系统，目标测量执行子系统是至少一个测量执行子系统102中被调度到的测量执行子系统，目标测量执行子系统是与待测路径适配或对应的测量执行子系统，例如可以是距离待测路径最近的，或者是部署在待测路径上的，或者是执行性能较佳的。目标测量执行子系统具体用于：在被调度到的时候，根据调度子系统101下发的测量规则，以源端设备的名义生成测量请求报文，并将该测量请求报文注入到待测路径上，以使待测路径上的至少部分网元设备对该测量请求报文进行转发并产生测量记录信息。如图1b所示，目标测量执行子 系统将该测量请求报文注入到待测路径上具体是指，将测量请求报文注入到待测路径上的某个网元设备内，以便测量请求报文自该网元设备开始朝向目的端设备不断被转发下去。在本实施例中，将待测路径上被注入测量请求报文的网元设备称为源网元设备，源网元设备可以是待测路径上的任一网元设备，具体可根据测量需求而定。
其中，测量请求报文是指以源端设备的名义发往目的端设备的报文，需要说明的是，该测量请求报文并非由源端设备生成，也并非从源端设备发出，而是由目标测量执行子系统代替源端设备生成并注入到待测路径上的，如图1b所示。在本实施例中，并不对测量请求报文的报文格式进行限定，凡是能够体现“以源端设备的名义发往目的端设备”的报文格式均适用于本申请实施例。下面是本申请实施例给出一种测量请求报文的标准化格式，可以支持各种通信协议，例如可以支持TCP或UDP协议等。如下所述，该测量请求报文至少包括：协议头部和测量头部；其中，协议头部可以是IP头部或UDP头部等，以用于支持不同的通信协议；关于各头部包括的字段信息如下：
IP头部：
dscp：IP优先和服务类型字段的组合，其取值可以是但不限于：63；
sip：源端设备的IP地址；
dip：目端设备的IP地址；
UDP头部：
dport：目的端口，例如可以是5000，但不限于此；
测量头部：
uuid：测量请求报文的唯一ID
tenant-id：源端设备所属租户的ID，该ID可表示租户的虚拟网络拓扑；
sip：源端设备的IP地址；
dip：目的端设备的IP地址；
mea_src_tenant_id：源网元设备所属网络空间的ID，一个租户可以有多个网络空间，不同网络空间可通过网络空间的ID进行区分；
mea src_ip：源网元设备的IP地址和/或承载源网元设备的物理机的IP地址；
timestamp：测量请求报文被源网元设备发送时的时间戳，为视区分，称为第一时间戳；
mea_reply_tenant_id：用于在测量回复报文中承载返回测量回复报文的网元设备所属租户的ID，在测量请求报文中保留用，缺省为0；
mea reply_ip：用于在测量回复报文中承载返回测量回复报文的网元设备的IP地址，在测量请求报文中保留用；
reply-timestamp：用于在测量回复报文中承载返回测量回复报文的时间戳，在测量请求报文中保留用。
基于上述报文格式，目标测量执行子系统在生成测量请求报文时，具体可以根据源端设备和目的端设备的信息，生成协议头部，若是IP头部，则根据源端设备和目的端设备的IP地址生成IP头部；若是UDP头部，则根据源端设备和目的端设备的端口号和协议类型等信息生成UDP头部；进而，根据目标租户的ID(对应tenant-id字段)、源网元设备的详情信息以及发出测量请求报文的第一时间戳(对应timestamp字段)，生成测量请求报文的测量头部；根据协议头部和测量头部，生成测量请求报文。其中，源网元设备的详情信息包括但不限于：源网元设备所属网络空间的ID(对应mea_src_tenant_id字段)、源网元设备的IP地址和/或承载源网元设备的物理机的IP地址(对应mea src_ip字段)。其中，对目标测量执行子系统来说，若其并未部署在源网元设备内，则可以将其向源网元设备注入测量请求报文的时间作为第一时间戳；若其部署在源网元设备内，则可以获知源网元设备发出测量请求报文的时间作为第一时间戳。
在此说明，上述源网元设备是指待测路径上测量请求报文被注入到的网元设备。可选地，根据目标测量执行子系统与源网元设备之间的部署关系，目标测量执行子系统向源网元设备注入测量请求报文的方式会有所不同。例如，在一可选实施例中，可以在云网络中各网元设备旁边旁挂测量执行子系统，基于此，目标测量执行子系统生成测量请求报文之后，可以将测量请求报文发送给与其对应的目标测量执行子系统。又例如，在另一可选实施例中，可以直接在云网络中各网元设备中部署测量执行子系统，基于此，目标测量执行子系统向源网元设备注入测量请求报文的过程实际上是部署于源网元设备上的目标测量执行子系统生成测量请求报文的过程，在该情况下，测量请求报文的生成时间与测量请求报文的发送时间基本相同或一致。
在本实施例中，无论测量执行子系统采用何种部署方式，均不对源网元设备做具体限定，源网元设备可以是待测路径上的任一网元设备。在本实施例中，待测路径除了包含源端设备和目的端设备之外，还包括连接于源端设备和目的端设备之间的至少一个网元设备，则目标测量执行子系统可以将测量请求报文发送至待测路径上的任何一个网元设备，即源网元设备，由源网元设备将测量请求报文转发出去，以使测量请求报文继续被待测路径上的其它网元设备进行转发。需要说明的是，测量请求报文具体被注入到哪 个网元设备，又被转发至哪个网元设备，可根据网络质量测量需求而定。与源网元设备对应，将测量请求报文最终被转发到的网元设备称为目标网元设备。
在一可选实施例中，如果需要测量局部路径的网络质量，则可以将测量请求报文注入到该局部路径的起始网元设备，由该起始网元设备将测量请求报文转发出去，测量请求报文经后续网元设备继续转发直至该局部路径的结束网元设备为止。在该可选实施例中，上文中的源网元设备即为起始网元设备，相应地，目标网元设备即为结束网元设备。其中，局部路径是指待测路径上的一段路径。其中，目标测量执行子系统可以限定测量请求报文的转发次数，通过转发次数确定该测量请求报文到达结束网元设备之后不会被继续转发。或者，目标测量执行子系统也可以向结束网元设备发送终止指令，以指示结束网元设备不要继续向外转发测量请求报文。
或者，
在另一可选实施例中，目标测量执行子系统可以将测量请求报文发送给待测路径上与源端设备直接连接的第一网元设备，由第一网元设备将测量请求报文转发出去，以使测量请求报文继续在待测路径上被转发直至第二网元设备，第二网元设备是待测路径上与目的端设备直接连接的网元设备。在图1b中，以将测量请求报文注入到第一网元设备，并被逐步转发至第二网元设备为例进行图示。在该可选实施例中，上文中的源网元设备即为第一网元设备，相应地，目标网元设备即为第二网元设备。在该实施例中，待测路径上每个中间的网元设备都会进行测量请求报文的转发，都可以生成相应的测量记录信息，不仅可以测量整个路径的网络质量，还可以对中间经过的各网元设备的网络质量进行测量，具备快速且准确定位网络问题的能力。
进一步可选地，在上述实施例中，第二网元设备收到测量请求报文之后，不再将测量请求报文转发出去，而是在生成测量记录信息之后，将测量请求报文丢弃，这样可以降低测量请求报文对目的端设备的侵入，做到租户对网络测量的无感知。在图1b中，以第二网元设备将测量请求报文丢弃，不再向目的端设备转发为例进行图示。除此之外，如果需要测量第二网元设备到目的端设备之间的网络质量，或者需要测量目的端设备的网络质量，则第二网元设备也可以将测量请求报文转发给目的端设备，目的端设备采用与前面网元设备相同的方式生成测量记录信息。进一步，为了提高租户的体验度，目的端设备是否需要进行网络质量测量，可由目的端设备所属的租户确定，该租户可以选择开启目的端设备的网络质量测量功能，则第二网元设备可以继续将测量请求报文转发至目的端设备；如果该租户并未开启目的端设备的网络质量测量功能，则第二网元设备会 直接丢弃测量请求报文，不会继续往外转发。
在本实施例中，在源网元设备将测量请求报文转发出去之后，待测路径上收到测量请求报文的其它网元设备可以向源网元设备返回测量回复报文。在本实施例中，并不限定测量回复报文的报文格式，与测量请求报文类似，本申请实施例提供一种测量回复报文的标准化格式，如下所述，该测量回复报文至少包括：协议头部和测量头部；其中，协议头部可以是IP头部或UDP头部等，以支持不同通信协议，关于各头部包括的字段信息如下：
IP头部：
dscp：IP优先和服务类型字段的组合，其取值可以是但不限于：62
sip：目端设备的IP地址；
dip：源端设备的IP地址；
UDP头部：
dport：目的端口，同测量请求报文，其取值可以直接复制测量请求报文的中UDP头部的取值；
测量头部：
其中，测量回复报文中测量头部与包含的字段信息与测量请求报文中测量头部包含的字段信息相同，可以直接复制测量请求报文中测量头部中的字段取值，但是补充mea_reply_tenant_id、mea_reply_ip、reply-timestamp这几个专用于测量回复报文的字段值。基于此，可知测量回复报文中测量头部包含的字段信息如下：
uuid：测量回复报文的唯一ID，可以与测量请求报文中的取值相同或者对应；
tenant-id：源端设备所属租户的ID；
sip：源端设备的IP地址；
dip：目的端设备的IP地址；
mea_src_tenant_id：源网元设备所属网络空间的ID；
mea src_ip：源网元设备的IP地址和/或承载源网元设备的物理机的IP地址；
timestamp：测量请求报文被发送时的时间戳，即第一时间戳；
mea_reply_tenant_id：返回测量回复报文的网元设备所属租户的ID；
mea reply_ip：返回测量回复报文的网元设备的IP地址；
reply-timestamp：返回回复测量报文的时间戳。
基于上述测量回复报文的格式，对于需要返回测量回复报文的网元设备来说，可以 根据源端设备和目的端设备的信息，生成测量回复报文中的协议头部；根据网元自身所属租户的ID、网元自身的详情信息以及第一时间戳等信息，生成测量回复报文中的测量头部；根据协议头部和测量头部生成测量回复报文。需要说明的是，网元设备在生成测量回复报文时，有些信息可以直接从测量请求报文中拷贝；或者，也可以直接在测量请求报文的基础上对相关信息进行修改，从而得到测量回复报文。对源网元设备来说，除了将测量请求报文转发出去之外，还可以接收其它网元设备返回的测量回复报文。
在本实施例中，测量请求报文的作用是让测量请求报文经过的网元设备或端设备产生测量记录信息。其中，测量记录信息是测量请求报文经过的网元设备或端设备所记录的一些与网络质量测量相关的信息，例如可以是租户的信息、网元设备的信息以及测量请求报文的收发时间等信息。其中，租户的信息可用于分析待测路径涉及了哪些租户，是否跨租户；网元设备的信息可用于分析测量请求报文实际经过了哪些网元设备，进一步结合测量请求报文应该经过的网元设备，可以分析待测路径、租户网络或整个云网络的丢包情况；测量请求报文的收发时间可用于分析网元设备内部的网络延时，租户网络或者整条路径的网络延时，进一步还可以分析租户网络或整个云网络的网络延时等。
在本申请可选实施例中，将测量记录信息分为两种类型，一种是路径记录信息，主要用于记录测量请求报文经过的路径信息；一种是延时记录信息，主要用于记录测量请求报文在转过过程中的传输延时。对于每个发送测量请求报文的网元设备均可以生成路径记录信息；对于延时记录信息可由源网元设备，即上述起始网元设备或第一网元设备负责生成，可根据测量请求报文和测量回复报文相关的信息生成测量请求报文的延时记录信息。在本申请实施例中，并不限定路径记录信息和延时记录信息的信息格式，本实施例给出的一种路径记录信息和延时记录信息的示例性格式如下：
路径记录信息：
Tenant-id：生成该路径记录信息的网元设备所属租户的ID，该租户ID与测量请求报文中的租户ID可能是不同的；
GEN-IP：生成该路径记录信息的网元设备的IP地址和/或承载该网元设备的物理机的IP地址；其中，网元设备的IP地址与物理机的IP地址可能相同，也可能不相同；
GEN-POINT：保留字段，对此不做限定；
GEN-TS：记录生成路径记录信息的时间戳，为便于区分，称为第二时间戳/或第三时间戳，关于第二时间戳或第三时间戳的解释参见下文；
GEN-DROP-CODE：如果因为丢弃数据包，填写该丢包原因，其中，丢包原因可以是硬件原因、限速原因等；
是否是目标网元设备：是一个布尔型(bool)变量，其取值为是或不是；
下一跳网元IP：下一跳网元设备的IP地址；
测量头部：同测量请求报文中的测量头部，其具体取值可直接复制测量请求报文中测量头部的取值。
需要说明的是，上述路径记录信息包含的各字段可根据应用需求灵活调整，可以增加或减少，并不限于这些字段。基于此，源网元设备(即第一网元设备或起始网元设备)和接收到测量请求报文的其它网元设备可以记录网元自身所属租户的ID、网元自身的详情信息、接收到测量请求报文的第二时间戳、转发测量请求报文的第三时间戳中的至少一种信息，以生成路径记录信息。
在一可选实施例中，收到测量请求报文的其它网元设备可以生成一个路径记录信息，则GEN-TS字段可以包括两个时间戳，即第二时间戳和第三时间戳；第二时间戳表示接收到测量请求报文的时间，第三时间戳表示将测量请求报文转发出去的时间。或者，在另一可选实施例中，收到测量请求报文的其它网元设备可以生成两个路径记录信息，即在接收到测量请求报文时，生成一个路径记录信息，则该路径记录信息中的GEN-TS字段记录第二时间戳；在将测量请求报文转发出去时，再次生成一个路径记录信息，则该路径记录信息中的GEN-TS字段记录第三时间戳。
延时记录信息：
Tenant-id：生成延时记录信息的网元设备(即源网元设备)所属租户的ID，在源网元设备生成延时记录信息的情况下，该租户ID与测量请求报文中的租户ID相同；
GEN-IP：生成延时记录信息的网元设备(即源网元设备)的IP地址和/或承载该网元设备的物理机的IP地址；其中，这两个IP地址可能相同，也可能不相同；
GEN-POINT：保留用字段，对此不做限定；
GEN-TS：记录发送测量请求报文的时间戳，即第一时间戳；
网元IP：产生测量回复报文的网元设备的IP地址；
测量头部：同测量回复报文中的测量头部，其取值可直接复制测量回复报文中测量头部的取值，但是源网元设备收到测量回复报文后会修改测量头部的reply–timesatmp字段，将其改为接收到测量回复请求报文的时间戳，为视区别，称为第四时间戳。
基于上述延时记录信息的格式，源网元设备可以根据发送测量请求报文的第一时间戳和接收到测量回复报文的第四时间戳，生成延时记录信息。其中，源网元设备可以从测量回复报文中读取第一时间戳，将第一时间戳记录到延时记录信息中；另外，记录接收到测量回复报文的时间戳作为第四时间戳，并将第四时间戳记录到延时记录信息中。
上述源网元设备和其它网元设备在生成路径记录信息或延时记录信息之后可上报给测量分析子系统103；在路径记录信息和延时记录信息的基础上，测量分析子系统103可在路径、租户以及云网络中至少一个维度上进行网络质量分析。具体地，测量分析子系统103可以根据收集到的延时记录信息和路径记录信息，分析待测路径的网络延时、待测路径上的丢包率、目标租户的网络延时、目标租户的丢包率、整个云网络的网络延时以及整个云网络的丢包率中的至少一个。
如图1c所示，测量分析子系统103的一种内部实现结构包括：路径分析模块103a、丢包统计模块103b、延时计算模块103c和结果聚合模块103d。
其中，源网元设备或其它网元设备生成的路径记录信息被送入路径分析模块103a，路径分析模块103a根据路径记录信息中测量头部中的报文ID，统计出同一待测路径上的路径记录信息，即包含相同报文ID的路径记录信息即为同一待测路径上的路径记录信息，从路径维度对路径记录信息进行分类，并将分类得到的不同待测路径上的路径记录信息分别提供给丢包统计模块130b。丢包统计模块103b针对每个待测路径上的路径记录信息，结合路径记录信息中的租户ID和网元设备的IP，可以分析出测量请求报文经过了哪些网元设备，进一步结合路径记录信息中的丢包原因还可以分析出哪些网元设备发生了丢包以及丢包原因是什么；进一步根据这些信息，可以从路径维度统计出不同待测路径上的丢包率；更进一步，结合租户ID还可以统计同一租户对应的待测路径，根据同一租户对应的待测路径上的丢包率，从租户维度统计该租户网络的丢包率；更进一步，还可以从云网络维度，根据不同租户网络的丢包率统计出整个云网络的丢包率。丢包统计模块130b将统计出的各维度上的丢包率提供给结果聚合模块103d。
源网元设备生成的延时记录信息被送入延时计算模块103c；延时记录模块103c可以根据延时记录信息中测量头部中的报文ID，统计出同一待测路径上的延时记录信息；进一步，根据延时记录信息中的第一时间戳和第二时间戳，可以测量出源网元设备到不同网元设备之间的网络延时；更进一步，还可以从路径维度，统计出整个待测路径的网络延时；更进一步，还可以从云网络维度，根据不同待测路径的网络延时统计出整个云网络的网络延时。延时计算模块103c将统计出的各维度上的丢包率提供给结果聚合模块 103d。
进一步可选地，路径记录信息也可以被送入延时计算模块103c；延时计算模块103c可以结合路径记录信息中的报文ID、网元IP等信息，识别出同一网元设备在接收测量请求报文时和转发测量请求报文时产生的路径记录信息；根据这两个路径记录信息中记录的第二时间戳和第三时间戳，计算出网元设备内部的处理延时，并将网元设备内部的处理延时提供给结果聚合模块103d。
结果聚合模块103d可以汇总各维度上的丢包率和各维度上的网络延时，按照网络质量测量需求进行网络质量分析，并输出网络质量分析结果。可选地，如图1a所示，网络质量分析结果还可以输出给调度子系统101，以影响调度子系统101意图感知，形成一测量闭环系统，提高网络质量测量效果。
采用本申请实施例的测量系统，通过感知租户的测量意图，基于测量意图进行网络质量测量，可以将租户从网络质量测量中解放出来，相比于租户主动测量网络质量的方案，本申请实施例避免了租户对复杂的测量规则的管理，简化了网络质量测量工作；采用旁路注包的方式，无需在租户的网络环境中部署测量服务，有利于降低了对租户网络环境的侵入；另外，本实施例中的旁路注包方式是主动注包，可避免对租户实际应用流量的依赖，即使在租户没有产生实际应用流量的情况下也可以对网络质量进行测量。再者，本申请实施例给出了标准化的测量请求报文、测量回复报文以及路径记录信息和延时记录信息，基于这些标准化的报文或信息格式可实现一种标准化的网络质量测量协议，这些标准化报文和信息能够不仅可以承载虚拟网元的信息，还可以同时承载物理机和租户的信息，从而实现对物理网络、Overlay网络和租户网络三层网络的网络质量测量，可以支持网络端到端以及中间节点质量测量的全覆盖，具备快速定位网络问题的能力。
在上述实施例中，并不限定测量系统的部署实施方式，在一可选实施例中，测量系统可独立部署在云网络之外，并与云网络中的网元设备进行通信连接，如图1d所示。在另一可选实施例中，测量系统中的各子系统可分布部署在云网络中实现。鉴于此，本申请实施例还提供一种具有网络质量测量功能的云网络，如图2a所示，该云网络200包括：物理网络201，以及部署于物理网络201之间上的虚拟网络202。进一步，如图2a所示，虚拟网络202包括多租户网络2021。在图2a中，以租户1、租户2和租户3的网络为例进行图示，但并不限于此。
其中，物理网络201包括服务器、机柜、路由器、交换机等物理机201a，以及用于 实现这些物理机之间的网络连接的物理连接线201b，例如同轴线缆、网线、光纤等。在图2a中，以物理机201a包括物理服务器S1、物理服务器S2、物理服务器S3以及连接于这些物理服务器之间的物理交换机和物理路由器为例进行图示，但本领域技术人员应该理解整个物理网络201包含的网络资源以及网络架构并不局限于此。
在本申请实施例中，采用虚拟化技术对物理网络201中的网络资源进行虚拟化，从而得到承载于物理网络201之上的虚拟网络202。如图2a所示，该虚拟网络202包括多租户网络2021，在图2a中，以租户1、租户2和租户3的网络为例进行图示，但并不限于此。进一步可选地，租户1、租户2和租户3的网络可以实现为各自的VPC，但不限于此。不同租户的网络环境之间相互隔离，每个租户网络包括租户可见的端设备202a，这些端设备202a承载于物理网络201中的物理机201a上，在图2a中，以端设备202a是VM为例进行图示，这些VM位于物理网络201中的物理服务器S1和S2上。进一步，如图2a所示，虚拟网络202中还包括：用于在不同端设备202a之间进行流量转发和网络互联的网元设备202b，网元设备202b可以是虚拟交换机或虚拟网关，在图2a中以网元设备202b是虚拟交换机或虚拟网关为例进行图示，但并不限于此。其中，网元设备202b属于虚拟网络202，但不属于任何租户网络。
在一可选实施例中，如图2b所示，采用Overlay技术在虚拟网络202中实现Overlay网络2022，Overlay网络2022介于物理网络201和多租户网络2021之间，上述网元设备202b可位于Overlay网络2022中实现。Overlay网络2022是多租户网络2021与物理网络201之间的桥梁，使得多租户网络2021中端设备202a之间的互联跳出物理网络的束缚，为最终实现网络资源的灵活定义、按需分配、按需调整提供了条件。这样整个云网络200可实现为包括物理网络201、Overlay网络2022和多租户网络2021的三层网络架构，但对云网络架构的划分并不限于此。
进一步，如图2a或图2b所示，在本实施例的云网络200中，还包括：调度设备203和分析设备204。在本实施例中，并不限定调度设备203和分析设备204在云网络200中的部署位置。在一可选实施例中，调度设备203和分析设备204部署在物理网络201中实现，例如两者部署在物理网络201中同一台物理机201a上实现，或者，两者分别部署在物理网络201中不同物理机201a上实现；其中，可以在物理网络201中新增一台或多台物理机201a，用于部署调度设备203和分析设备204；或者，也可以将调度设备203和分析设备204直接部署到物理网络201中原有的物理机201a上。在另一可选实施例中，调度设备203和分析设备204部署在虚拟网络202中除多租户网络2021之外的其他网络 位置上实现，例如两者可以部署在Overlay网络2022中实现。在图2a或图2b中，以调度设备203和分析设备204均部署在物理网络201中物理服务器S3上实现为例进行图示，但并不限于此。另外，虚拟网络202中的各个网元设备202b除了具有流量转发、报文处理以及网络互联等功能之外，还可以配合调度设备203和分析设备204实现网络质量的测量，主要是指进行测量请求报文的注入、转发、测量记录信息的生成和上报等处理，根据网元设备202b在待测路径上的位置不同，各网元设备202b的处理操作会有所差异。本实施例下面将重点对网元设备202b配合调度设备203和分析设备204实现网络质量测量的过程进行详细描述。
在本实施例中，调度设备203感知目标租户的测量意图，生成与测量意图适配的测量规则并下发给源网元设备，所述测量规则包括待测路径上的源端设备和目的端设备，源网元设备是待测路径上的任一网元设备202b。关于调度设备203感知目标租户测量意图和生成测量规则的详细实施方式，可参见前述实施例中调度子系统101感知测量意图和生成测量规则的详细实施方式，在此不再赘述。
在本实施例中，源网元设备负责根据测量规则生成测量请求报文，并将测量请求报文转发出去，所述测量请求报文用于供源网元设备和待测路径上收到测量请求报文的其它网元设备生成测量记录信息。源网元设备可以是待测路径上的任一网元设备，具体可根据测量需求而定。在一可选实施例中，源网元设备是待测路径上与源端设备直接连接的第一网元设备，在该情况下，测量请求报文自第一网元设备开始被转发至目标网元设备。同理，本申请实施例也不对目标网元设备进行限定，可选地，目标网元设备是待测路径上与目的端设备直接连接的第二网元设备。
在一可选实施例中，源网元设备生成测量请求报文的实施方式包括：根据源端设备和目的端设备的信息，生成测量请求报文中的协议头部；根据目标租户的标识、源网元设备的详情信息以及第一时间戳，生成测量请求报文中的测量头部；根据协议头部和测量头部，生成测量请求报文；其中，第一时间戳表示源网元设备对外发送测量请求报文的时间。
其中，源网元设备或其它网元设备除了生成测量请求报文并将测量请求报文发送出去之外，还会生成测量记录信息。进一步可选地，测量记录信息包括路径记录信息。基于此，源网元设备或其它网元设备生成测量记录信息，包括：记录网元自身所属租户的标识、网元自身的详情信息、接收到测量请求报文的第二时间戳、转发测量请求报文的第三时间戳中的至少一种信息，以生成路径记录信息。其中，路径记录信息被源网元设 备或其它网元设备上报给分析设备204。
进一步，其它网元设备在接收到测量请求报文后，还可以生成测量回复报文并返回给源网元设备，在测量回复报文中包括第一时间戳，第一时间戳来自测量请求报文。相应地，源网元设备还用于：根据测量回复报文中携带的第一时间戳和接收到测量回复报文的第四时间戳，生成延时记录信息。其中，延时记录信息被源网元设备上报给分析设备204。
关于生成测量请求报文以及生成路径记录信息和延时记录信息的详细过程，可参见前述实施例的描述，在此不再赘述。
在本实施例中，分析设备204用于根据源网元设备和其它网元设备生成的测量记录信息进行网络质量分析。关于分析设备204根据测量记录信息进行网络质量分析的详细实施方式，可参见前述实施例中测量分析子系统103进行网络质量分析的详细实施方式，在此不再赘述。
在本实施例的云网络中，增设调度设备和分析设备，并对虚拟网络中的网元设备进行功能拓展，使得该云网络可以通过感知租户的测量意图，基于测量意图进行网络质量测量，一方面可以将租户从网络质量测量中解放出来，简化网络质量测量工作；另一方面，采用旁路注包的方式，无需在租户的网络环境中部署测量服务，有利于降低了对租户网络环境的侵入；另外，采用主动注包方式，可避免对租户实际应用流量的依赖，即使在租户没有产生实际应用流量的情况下也可以对网络质量进行测量。再者，结合本申请实施例给出的标准化的报文格式和信息格式，云网络中的报文和信息能够同时承载虚拟网元、物理机和租户的信息，可以实现对物理网络、Overlay网络和租户网络三层网络的网络质量测量，可以支持网络端到端以及中间节点质量全覆盖，具备快速定位网络问题的能力。
在此说明，本实施例的云网络支持多租户，对租户数量不做限定，且支持租户部署多种应用；另外，本实施例也不限定云网络的链路长度，可根据应用需求灵活部署、调整，而且本实施例的云网络具有易于部署的优点，可以根据应用需求灵活、广泛部署。
进一步，本申请实施例还提供一种网络质量测量方法，该方法适用于上述实施例提供的测量系统或云网络，但并不限于此。如图3a所示，该方法包括：
31a、感知云网络中目标用户的测量意图，并生成与该测量意图适配的测量规则，该测量规则包括待测路径上的源端设备和目的端设备。
32a、根据上述测量规则生成测量请求报文，并将测量请求报文注入到待测路径上， 以使待测路径上的至少部分网元设备生成测量记录信息。
33a、根据上述至少部分网元设备生成的测量记录信息进行网络质量分析。
在一可选实施例中，上述感知云网络中目标用户的测量意图，包括：根据目标租户提交的应用需求变化信息，生成目标租户的网络拓扑变化信息；根据网络拓扑变化信息，确定目标租户的测量意图。
在一可选实施例中，上述生成与测量意图适配的测量规则，包括：根据目标租户的测量意图和目标租户的网络配置信息，确定待测路径上的源端设备和目的端设；根据待测路径上的源端设备和目的端设备，生成测量规则。
在一可选实施例中，上述根据目标租户的测量意图和目标租户的网络配置信息，确定待测路径上的源端设备和目的端设，包括：从目标租户的测量意图中，解析出待测量的源端设备；根据源端设备所在的网络拓扑，结合目标租户的网络配置信息，获取与源端设备存在访问关系的潜在路径；从潜在路径上的另一端设备中，确定与源端设备对应的目的端设备。
在一可选实施例中，上述从潜在路径上的另一端设备中，确定与源端设备对应的目的端设备，包括：若目标租户选择全覆盖测量方式，则将全部潜在路径上的另一端设备均作为与源端设备对应的目的端设备；若目标租户选择选择性测量方式，则从全部潜在路径中选择部分潜在路径，将部分潜在路径上的另一端设备作为与源端设备对应的目的端设备。
进一步可选地，本实施例的方法还包括：向目标租户展示人机交互界面，人机交互界面上设置有测量方式选项，以供目标租户选择；响应于目标租户针对测量方式选项发起的选择操作，确定目标租户选择的测量方式，目标租户选择的测量方式为全覆盖测量方式或选择性测量方式。
在一可选实施例中，上述根据测量规则生成测量请求报文，包括：根据源端设备和目的端设备的信息，生成测量请求报文中的协议头部；根据目标租户的标识、第一网元设备的详情信息以及第一时间戳，生成测量请求报文中的测量头部；根据协议头部和测量头部，生成测量请求报文；其中，第一时间戳表示发出测量请求报文的时间。
在一可选实施例中，上述将测量请求报文注入到待测路径上，以使待测路径上的至少部分网元设备生成测量记录信息，包括：将测量请求报文发送给待测路径上的源网元设备，自源网元设备开始向待测路径上的其它网元设备转发测量请求报文，以使源网元设备和其它网元设备生成测量记录信息。
其中，根据方法执行主体与源网元设备的部署关系，向源网元设备注入测量请求报文的方式会有所不同。在一可选实施例中，可以在云网络中各网元设备旁边旁挂测量执行子系统，由测量执行子系统生成测量请求报文并注入到与其存在旁挂关系的网元设备中；基于此，向源网元设备注入测量请求报文具体是指与源网元设备存在旁挂关系的测量执行子系统在生成测量报文之后，将测量请求报文发送给源网元设备的过程。在另一可选实施例中，可以对云网络中各网元设备进行功能拓展，让各网元设备具有根据测量规则生成测量请求报文的功能；基于此，向源网元设备注入测量请求报文具体是指源网元设备基于拓展的报文生成功能生成测量请求报文的过程。
在一可选实施例中，源网元设备或其它网元设备生成测量记录信息，包括：记录网元自身所属租户的标识、网元自身的详情信息、接收到所述测量请求报文的第二时间戳、转发所述测量请求报文的第三时间戳中的至少一种信息，以生成路径记录信息。
在一可选实施例中，源网元设备生成测量记录信息还包括：接收其它网元设备发送的测量回复报文，所述测量回复报文包括所述第一时间戳；根据所述测量回复报文中的第一时间戳和接收到测量回复报文的第四时间戳，生成延时记录信息。
在一可选实施例中，上述根据至少部分网元设备生成的测量记录信息进行网络质量分析，包括：根据上述延时记录信息和路径记录信息，在待测路径、目标租户以及云网络中的至少一个维度上分析网络延时和丢包率。
除了图3a所示网络质量测量方法之外，本申请实施例还提供一种测量规则生成方法，该方法是从调度子系统或调度设备的角度进行的描述，如图3b所示，该方法包括：
31b、感知云网络中目标租户的测量意图；
32b、生成与测量意图适配的测量规则，所述测量规则包括待测路径上的源端设备和目的端设备；
33b、将测量规则下发给待测路径上的源网元设备，以使所述源网元设备生成测量请求报文并转发出去，所述测量请求报文用于供源网元设备和待测路径上收到测量请求报文的其它网元设备生成测量记录信息以进行网络质量分析。
可选地，源网元设备可以是待测路径上的任一网元设备，例如可以是待测路径上与源端设备直接连接的第一网元设备。
进一步，本申请实施例还提供一种报文传输方法，该方法主要是从云网络中网元设备的角度进行的描述，如图3c所示，该方法包括：
31c、源网元设备接收测量规则，所述测量规则包括待测路径上的源端设备和目的端 设备，源网元设备是待测路径上的网元设备；
32c、根据测量规则生成测量请求报文，并将测量请求报文转发出去，所述测量请求报文用于供源网元设备和待测路径上收到所述测量请求报文的其它网元设备生成测量记录信息以进行网络质量分析。
在本实施例中，并不限定测量规则的来源和生成方式，可选地，该测量规则可以采用上述实施例中的方式生成。
关于上述方法实施例中各步骤的详细描述和解释，可参见前述系统实施例，在此不再赘述。
需要说明的是，上述实施例所提供方法的各步骤的执行主体均可以是同一设备，或者，该方法也由不同设备作为执行主体。比如，步骤31a至步骤33a的执行主体可以为设备A；又比如，步骤31a执行主体可以为设备A，步骤32a的执行主体可以为设备B；等等。
另外，在上述实施例及附图中的描述的一些流程中，包含了按照特定顺序出现的多个操作，但是应该清楚了解，这些操作可以不按照其在本文中出现的顺序来执行或并行执行，操作的序号如31a、32a等，仅仅是用于区分开各个不同的操作，序号本身不代表任何的执行顺序。另外，这些流程可以包括更多或更少的操作，并且这些操作可以按顺序执行或并行执行。需要说明的是，本文中的“第一”、“第二”等描述，是用于区分不同的消息、设备、模块等，不代表先后顺序，也不限定“第一”和“第二”是不同的类型。
图4为本申请示例性实施例提供的一种网络质量测量装置的结构示意图。如图4所示，该装置包括：意图感知模块41、规则生成模块42、报文生成模块43、报文注入模块44和质量分析模块45。
意图感知模块41，用于感知云网络中目标租户的测量意图。规则生成模块42，用于生成与意图感知模块41感知到的测量意图适配的测量规则，所述测量规则包括待测路径上的源端设备和目的端设备。报文生成模块43，用于根据规则生成模块42所生成的测量规则生成测量请求报文。报文注入模块44，用于将报文生成模块43生成的测量请求报文注入待测路径上，以使待测路径上的至少部分网元设备生成测量记录信息。质量分析模块45，用于根据至少部分网元设备生成的测量记录信息进行网络质量分析。
在一可选实施例中，意图感知模块41具体用于：根据目标租户提交的应用需求变化信息，生成所述目标租户的网络拓扑变化信息；根据所述网络拓扑变化信息，确定所述 目标租户的测量意图。
在一可选实施例中，规则生成模块42具体用于：根据目标租户的测量意图和目标租户的网络配置信息，确定待测路径上的源端设备和目的端设；根据待测路径上的源端设备和目的端设备，生成测量规则。
进一步可选地，规则生成模块42在确定待测路径上的源端设备和目的端设备时，具体用于：从目标租户的测量意图中，解析出待测量的源端设备；根据源端设备所在的网络拓扑，结合目标租户的网络配置信息，获取与源端设备存在访问关系的潜在路径；从潜在路径上的另一端设备中，确定与源端设备对应的目的端设备。
更进一步，规则生成模块42在确定与源端设备对应的目的端设备时，具体用于：若目标租户选择全覆盖测量方式，则将全部潜在路径上的另一端设备均作为与源端设备对应的目的端设备；若目标租户选择选择性测量方式，则从全部潜在路径中选择部分潜在路径，将所述部分潜在路径上的另一端设备作为与源端设备对应的目的端设备。
在一可选实施例中，如图4所示，该装置还包括：配置模块46，用于向目标租户展示人机交互界面，所述人机交互界面上设置有测量方式选项，以供所述目标租户选择；响应于目标租户针对测量方式选项发起的选择操作，确定所述目标租户选择的测量方式，所述目标租户选择的测量方式为全覆盖测量方式或选择性测量方式。
在一可选实施例中，报文注入模块44具体用于：将测量请求报文发送给待测路径上的源网元设备，自源网元设备开始向待测路径上的其它网元设备转发测量请求报文，以使源网元设备和其它网元设备生成测量记录信息。
在一可选实施例中，报文生成模块43具体用于：根据所述源端设备和目的端设备的信息，生成所述测量请求报文中的协议头部；根据所述目标租户的标识、源网元设备的详情信息以及第一时间戳，生成所述测量请求报文中的测量头部；根据所述协议头部和测量头部，生成所述测量请求报文，所述第一时间戳表示发出所述测量请求报文的时间。
在此说明，本实施例的网络质量测量装置可以分布部署在云网络中实现，具体地，意图感知模块41、规则生成模块42和质量分析模块45可以独立部署在云网络中，而报文生成模块43和报文注入模块44可以部署在云网络中的网元设备中实现。进一步可选地，在报文生成模块43和报文注入模块44部署在网元设备中实现的情况下，报文生成模块43和报文注入模块44可以实现为一个模块。
在报文生成模块43和报文注入模块44部署在源网元设备中实现的情况下，报文生成模块43或报文注入模块44还用于生成测量记录信息。具体地，报文生成模块43或报 文注入模块44记录网元自身所属租户的标识、网元自身的详情信息、接收到所述测量请求报文的第二时间戳、转发测量请求报文的第三时间戳中的至少一种信息，以生成路径记录信息。
进一步可选地，报文生成模块43或报文注入模块44还可以接收其它网元设备发送的测量回复报文，所述测量回复报文包括所述第一时间戳；根据测量回复报文中的第一时间戳和接收到测量回复报文的第四时间戳，生成延时记录信息。
在一可选实施例中，质量分析模块45具体用于：根据上述延时记录信息和路径记录信息，在待测路径、目标租户以及云网络中的至少一个维度上分析网络延时和丢包率。
以上描述了网络质量测量装置的内部功能和结构，如图5所示，实际中，该网络测量测量装置可实现为云计算设备，包括：存储器51、处理器52以及通信组件53。
存储器51，用于存储计算机程序，并可被配置为存储其它各种数据以支持在云计算设备上的操作。这些数据的示例包括用于在云计算设备上操作的任何应用程序或方法的指令，消息，图片，视频等。
处理器52，与存储器51耦合，用于执行存储器51中的计算机程序，以用于：感知云网络中目标用户的测量意图，并生成与所述测量意图适配的测量规则，所述测量规则包括待测路径上的源端设备和目的端设备；根据测量规则生成测量请求报文，并将测量请求报文注入到所述待测路径上，以使待测路径上的至少部分网元设备生成测量记录信息；根据至少部分网元设备生成的测量记录信息进行网络质量分析。
在一可选实施例中，处理器52在感知测量意图时，具体用于：根据目标租户提交的应用需求变化信息，生成所述目标租户的网络拓扑变化信息；根据所述网络拓扑变化信息，确定所述目标租户的测量意图。
在一可选实施例中，处理器52在生成测量规则时，具体用于：根据目标租户的测量意图和目标租户的网络配置信息，确定待测路径上的源端设备和目的端设；根据待测路径上的源端设备和目的端设备，生成测量规则。
进一步可选地，处理器52在确定待测路径上的源端设备和目的端设备时，具体用于：从目标租户的测量意图中，解析出待测量的源端设备；根据源端设备所在的网络拓扑，结合目标租户的网络配置信息，获取与源端设备存在访问关系的潜在路径；从潜在路径上的另一端设备中，确定与源端设备对应的目的端设备。
更进一步，处理器52在确定与源端设备对应的目的端设备时，具体用于：若目标租户选择全覆盖测量方式，则将全部潜在路径上的另一端设备均作为与源端设备对应的目 的端设备；若目标租户选择选择性测量方式，则从全部潜在路径中选择部分潜在路径，将所述部分潜在路径上的另一端设备作为与源端设备对应的目的端设备。
在一可选实施例中，处理器52还用于：向目标租户展示人机交互界面，所述人机交互界面上设置有测量方式选项，以供所述目标租户选择；响应于目标租户针对测量方式选项发起的选择操作，确定所述目标租户选择的测量方式，所述目标租户选择的测量方式为全覆盖测量方式或选择性测量方式。
在一可选实施例中，处理器52在将测量请求报文注入待测路径上时，具体用于：将测量请求报文发送给待测路径上的源网元设备，自源网元设备开始向待测路径上的其它网元设备转发测量请求报文，以使源网元设备和其它网元设备生成测量记录信息。
在一可选实施例中，处理器52在生成测量请求报文时，具体用于：根据所述源端设备和目的端设备的信息，生成测量请求报文中的协议头部；根据所述目标租户的标识、源网元设备的详情信息以及第一时间戳，生成所述测量请求报文中的测量头部；根据所述协议头部和测量头部，生成所述测量请求报文，所述第一时间戳表示发出所述测量请求报文的时间。
在此说明，本实施例的云计算设备可以分布部署在云网络中实现，具体地，关于测试报文生成和注入功能可部署在云网络中的网元设备中实现。基于此，处理器52还用于生成测量记录信息。处理器52具体用于：记录其所在网元自身所属租户的标识、网元自身的详情信息、接收到所述测量请求报文的第二时间戳、转发测量请求报文的第三时间戳中的至少一种信息，以生成路径记录信息。
进一步可选地，针对源网元设备，处理器52还用于：接收其它网元设备发送的测量回复报文，所述测量回复报文包括所述第一时间戳；根据测量回复报文中的第一时间戳和接收到测量回复报文的第四时间戳，生成延时记录信息。
在一可选实施例中，处理器52在进行网络质量分析时，具体用于：根据上述延时记录信息和路径记录信息，在待测路径、目标租户以及云网络中的至少一个维度上分析网络延时和丢包率。
进一步，如图5所示，云计算设备还包括：电源组件54等其它组件。图5中仅示意性给出部分组件，并不意味着云计算设备只包括图5所示组件。
相应地，本申请实施例还提供一种存储有计算机程序的计算机可读存储介质，当计算机程序被处理器执行时，致使处理器能够实现上述方法实施例中的各步骤。
相应地，本申请实施例还提供一种计算机程序产品，包括计算机程序/指令，当所述 计算机程序/指令被处理器执行时，致使处理器能够实现上述方法实施例中的各步骤。
上述实施例中的存储器可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
上述实施例中的通信组件被配置为便于通信组件所在设备和其他设备之间有线或无线方式的通信。通信组件所在设备可以接入基于通信标准的无线网络，如WiFi，2G、3G、4G/LTE、5G等移动通信网络，或它们的组合。在一个示例性实施例中，通信组件经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
上述实施例中的电源组件，为电源组件所在设备的各种组件提供电力。电源组件可以包括电源管理系统，一个或多个电源，及其他与为电源组件所在设备生成、管理和分配电力相关联的组件。
本领域内的技术人员应明白，本申请的实施例可提供为方法、系统、或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本申请是参照根据本申请实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或 多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。
以上所述仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的权利要求范围之内。

Claims (12)

  1. 一种云网络，其特征在于，包括：物理网络和承载于所述物理网络之上的虚拟网络；所述虚拟网络包括多租户网络，以及负责在所述多租户网络中不同端设备之间进行流量转发和互联的网元设备；所述云网络还包括：调度设备和分析设备；
    所述调度设备，用于感知目标租户的测量意图，生成与所述测量意图适配的测量规则并下发给源网元设备，所述测量规则包括待测路径上的源端设备和目的端设备，所述源网元设备是所述待测路径上的网元设备；
    所述源网元设备，用于根据所述测量规则生成测量请求报文，并将所述测量请求报文转发出去，所述测量请求报文用于供所述源网元设备和所述待测路径上收到所述测量请求报文的其它网元设备生成测量记录信息；
    所述分析设备，用于根据所述源网元设备和其它网元设备生成的测量记录信息进行网络质量分析。
  2. 根据权利要求1所述的云网络，其特征在于，所述源网元设备是所述待测路径上与所述源端设备直接连接的第一网元设备，所述测量请求报文自所述第一网元设备开始被转发至目标网元设备，所述目标网元设备是所述待测路径上与所述目的端设备直接连接的第二网元设备。
  3. 一种用于云网络的测量系统，其特征在于，包括：调度子系统、至少一个测量执行子系统和测量分析子系统；
    所述调度子系统，用于感知云网络中目标租户的测量意图，生成与所述测量意图适配的测量规则并下发给目标测量执行子系统，所述测量规则包括待测路径上的源端设备和目的端设备；所述目标测量执行子系统与所述待测路径对应；
    所述目标测量执行子系统，用于根据所述测量规则生成测量请求报文，并将所述测量请求报文注入到所述待测路径上，以使所述待测路径上的至少部分网元设备对所述测量请求报文进行转发并产生测量记录信息；
    所述测量分析子系统，用于根据所述至少部分网元设备产生的测量记录信息进行网络质量分析。
  4. 一种网络质量测量方法，其特征在于，包括：
    感知云网络中目标用户的测量意图，并生成与所述测量意图适配的测量规则，所述测量规则包括待测路径上的源端设备和目的端设备；
    根据所述测量规则生成测量请求报文，并将所述测量请求报文注入到所述待测路径 上，以使所述待测路径上的至少部分网元设备生成测量记录信息；
    根据所述至少部分网元设备生成的测量记录信息进行网络质量分析。
  5. 根据权利要求4所述的方法，其特征在于，感知云网络中目标用户的测量意图，包括：
    根据目标租户提交的应用需求变化信息，生成所述目标租户的网络拓扑变化信息；
    根据所述网络拓扑变化信息，确定所述目标租户的测量意图。
  6. 根据权利要求5所述的方法，其特征在于，生成与所述测量意图适配的测量规则，包括：
    根据所述目标租户的测量意图和所述目标租户的网络配置信息，确定待测路径上的源端设备和目的端设备；
    根据所述待测路径上的源端设备和目的端设备，生成测量规则。
  7. 根据权利要求6所述的方法，其特征在于，根据所述目标租户的测量意图和所述目标租户的网络配置信息，确定待测路径上的源端设备和目的端设备，包括：
    从所述目标租户的测量意图中，解析出待测量的源端设备；
    根据所述源端设备所在的网络拓扑，结合所述目标租户的网络配置信息，获取与所述源端设备存在访问关系的潜在路径；
    从所述潜在路径上的另一端设备中，确定与所述源端设备对应的目的端设备。
  8. 根据权利要求4-7任一项所述的方法，其特征在于，将所述测量请求报文注入到所述待测路径上，以使所述待测路径上的至少部分网元设备生成测量记录信息，包括：
    将所述测量请求报文发送给所述待测路径上的源网元设备，自所述源网元设备开始向所述待测路径上的其它网元设备转发所述测量请求报文，以使所述源网元设备和其它网元设备生成测量记录信息。
  9. 根据权利要求8所述的方法，其特征在于，根据所述测量规则生成测量请求报文，包括：
    根据所述源端设备和目的端设备的信息，生成所述测量请求报文中的协议头部；
    根据所述目标租户的标识、所述源网元设备的详情信息以及第一时间戳，生成所述测量请求报文中的测量头部；
    根据所述协议头部和测量头部，生成所述测量请求报文，所述第一时间戳表示发出所述测量请求报文的时间。
  10. 根据权利要求9所述的方法，其特征在于，所述源网元设备或所述其它网元设 备生成测量记录信息，包括：
    记录网元自身所属租户的标识、网元自身的详情信息、接收到所述测量请求报文的第二时间戳、转发所述测量请求报文的第三时间戳中的至少一种信息，以生成路径记录信息；所述源网元设备生成测量记录信息还包括：接收所述其它网元设备发送的测量回复报文，所述测量回复报文包括所述第一时间戳；根据所述测量回复报文中的第一时间戳和接收到所述测量回复报文的第四时间戳，生成延时记录信息；
    相应地，根据所述至少部分网元设备生成的测量记录信息进行网络质量分析，包括：根据所述延时记录信息和所述路径记录信息，在所述待测路径、所述目标租户以及所述云网络中的至少一个维度上分析网络延时和丢包率。
  11. 一种云计算设备，其特征在于，包括：存储器和处理器；所述存储器，用于存储计算机程序；所述处理器，用于执行所述计算机程序，以用于实现权利要求4-10任一项所述权利要求中的步骤。
  12. 一种存储有计算机程序的计算机可读存储介质，其特征在于，当所述计算机程序被处理器执行时，致使所述处理器能够实现权利要求4-10任一项所述权利要求中的步骤。
PCT/CN2022/090914 2021-05-07 2022-05-05 云网络、用于云网络的测量系统、方法、设备及存储介质 WO2022233297A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110496812.X 2021-05-07
CN202110496812.XA CN112994987B (zh) 2021-05-07 2021-05-07 云网络、用于云网络的测量系统、方法、设备及存储介质

Publications (1)

Publication Number Publication Date
WO2022233297A1 true WO2022233297A1 (zh) 2022-11-10

Family

ID=76337232

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/090914 WO2022233297A1 (zh) 2021-05-07 2022-05-05 云网络、用于云网络的测量系统、方法、设备及存储介质

Country Status (2)

Country Link
CN (1) CN112994987B (zh)
WO (1) WO2022233297A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112994987B (zh) * 2021-05-07 2021-09-14 阿里云计算有限公司 云网络、用于云网络的测量系统、方法、设备及存储介质
CN114726518B (zh) * 2022-03-31 2023-05-26 阿里云计算有限公司 用于云网络系统的通信方法、装置、系统及存储介质
CN115499317B (zh) * 2022-11-15 2023-04-07 阿里云计算有限公司 灰度验证方法、电子设备和可读存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102143022A (zh) * 2011-03-16 2011-08-03 北京邮电大学 用于ip网络的云测量装置和测量方法
CN103491129A (zh) * 2013-07-05 2014-01-01 华为技术有限公司 一种业务节点配置方法、业务节点池注册器及系统
US20190068495A1 (en) * 2017-08-22 2019-02-28 Cisco Technology, Inc. Cloud provider classification for different service deployment schemes
CN111225031A (zh) * 2019-12-17 2020-06-02 长沙星融元数据技术有限公司 云数据中心虚拟底层网络架构及其数据传输方法
CN111279315A (zh) * 2017-10-24 2020-06-12 思科技术公司 租户间工作负载性能相关联和推荐
US20200322249A1 (en) * 2019-04-05 2020-10-08 Google Llc Cloud Network Reachability Analysis
CN112532468A (zh) * 2019-09-19 2021-03-19 华为技术有限公司 网络测量系统、方法、设备及存储介质
CN112994987A (zh) * 2021-05-07 2021-06-18 阿里云计算有限公司 云网络、用于云网络的测量系统、方法、设备及存储介质

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10630557B2 (en) * 2015-10-19 2020-04-21 Nicira, Inc. Virtual network management
CN111884872B (zh) * 2020-07-01 2022-05-06 中国联合网络通信集团有限公司 一种业务路径的性能测试方法和装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102143022A (zh) * 2011-03-16 2011-08-03 北京邮电大学 用于ip网络的云测量装置和测量方法
CN103491129A (zh) * 2013-07-05 2014-01-01 华为技术有限公司 一种业务节点配置方法、业务节点池注册器及系统
US20190068495A1 (en) * 2017-08-22 2019-02-28 Cisco Technology, Inc. Cloud provider classification for different service deployment schemes
CN111279315A (zh) * 2017-10-24 2020-06-12 思科技术公司 租户间工作负载性能相关联和推荐
US20200322249A1 (en) * 2019-04-05 2020-10-08 Google Llc Cloud Network Reachability Analysis
CN112532468A (zh) * 2019-09-19 2021-03-19 华为技术有限公司 网络测量系统、方法、设备及存储介质
CN111225031A (zh) * 2019-12-17 2020-06-02 长沙星融元数据技术有限公司 云数据中心虚拟底层网络架构及其数据传输方法
CN112994987A (zh) * 2021-05-07 2021-06-18 阿里云计算有限公司 云网络、用于云网络的测量系统、方法、设备及存储介质

Also Published As

Publication number Publication date
CN112994987A (zh) 2021-06-18
CN112994987B (zh) 2021-09-14

Similar Documents

Publication Publication Date Title
WO2022233297A1 (zh) 云网络、用于云网络的测量系统、方法、设备及存储介质
US11196620B2 (en) Method and apparatus for NaaS device configuring service
CN106209490B (zh) 选择和监控多个服务关键性能指标的方法和系统
US11558426B2 (en) Connection tracking for container cluster
US10003540B2 (en) Flow forwarding method, device, and system
RU2643451C2 (ru) Система и способ виртуализации функции мобильной сети
KR101906742B1 (ko) Sdn 네트워크에 기초한 애플리케이션 서비스 체인에 대한 정책 및 과금 제어 방법 및 장치
US10397132B2 (en) System and method for granting virtualized network function life cycle management
US20160212066A1 (en) Software-Defined Information Centric Network (ICN)
WO2017031698A1 (zh) 一种获取vnf信息的方法、装置及系统
JP2017517170A (ja) Ｎｆｖシステムにおけるサービス実装のための方法および通信ユニット
US10630579B1 (en) Ensuring separate paths for network traffic between source devices and a destination device
EP4109251A1 (en) Vnf instantiation method and device
WO2024067338A1 (zh) 云组网系统、安全访问方法、设备及存储介质
KR20230009960A (ko) 네트워크 성능 측정 방법, 장치, 디바이스, 및 시스템, 및 저장 매체
US11381497B2 (en) Path selection method and device
US11646956B2 (en) Systems and methods for providing bidirectional forwarding detection with performance routing measurements
WO2023057794A1 (en) Method for aligning quality of service in mobile network and edge cloud
CN109450794B (zh) 一种基于sdn网络的通信方法及设备
WO2022253190A1 (zh) 业务流的性能检测方法、装置及通信网络
US11973843B2 (en) On demand end user monitoring for automated help desk support
WO2022253194A1 (zh) 报文转发方法、装置及通信网络
WO2022253192A1 (zh) 报文转发方法、装置及通信网络
WO2023011006A1 (zh) 一种通信方法、装置及设备
US10644994B1 (en) Ensuring separate network paths between pairs of source and destination devices

Legal Events

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

Ref document number: 22798637

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 18559434

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22798637

Country of ref document: EP

Kind code of ref document: A1