WO2007118396A1 - A method and system for network performance measurement - Google Patents
A method and system for network performance measurement Download PDFInfo
- Publication number
- WO2007118396A1 WO2007118396A1 PCT/CN2007/000503 CN2007000503W WO2007118396A1 WO 2007118396 A1 WO2007118396 A1 WO 2007118396A1 CN 2007000503 W CN2007000503 W CN 2007000503W WO 2007118396 A1 WO2007118396 A1 WO 2007118396A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- flow
- identifier
- measurement
- content
- message
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
- H04L41/5003—Managing SLA; Interaction between SLA and QoS
- H04L41/5009—Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
Definitions
- the present invention relates to measurement techniques in the field of communications, and more particularly to methods and systems for measuring network performance. Background technique
- IP network performance measurement methods There are two main types of IP network performance measurement methods: active measurement and passive measurement.
- Active measurement is the use of measurement tools to actively generate measurement traffic at selected measurement points, inject into the network, and analyze the performance of the network based on the measurement data flow.
- Passive measurement refers to the measurement of the network using measurement equipment on links or devices (such as routers, switches, etc.) without the need to generate excess traffic.
- Network performance measurements are divided into directions and can be divided into one-way network performance measurement and round-trip network performance measurement.
- Unidirectional network performance measurement is a measure of network performance in one direction from one measurement point A to another.
- the round-trip network performance measurement measures network performance indicators from one measurement point A to another measurement point B and back to measurement point A.
- the QoS may be different in both directions of the round-trip, or the QoS of the same path to and from the two directions is different, so the one-way network performance is not equal to the round-trip delay of the single ticket. Divide by two.
- network performance measurements are all one-way network performance measurements.
- the passive measurement system generally includes a measurement controller, a collector, and a measurement device (or a measurement point).
- the function of the measurement controller is as follows: 1. Configure the measurement points, including: The quintuple of the measured flow, the duration of the measurement. If you use sampling, you also need to configure the sampling algorithm for the measurement points.
- the function of the measuring device is as follows:
- Timestamp used to measure point delay.
- the delay between two measurement points is determined according to the timestamp information in the message summary data of different measurement points.
- the functions of the collector are as follows:
- the delay and loss of each ⁇ ⁇ ⁇ text can be calculated within a period of time (this The time period is called the calculation interval (Evaluation lnterval), and the network performance indicators such as average delay, maximum delay, delay, delay jitter, packet loss rate, and network unavailability.
- the flow identifier is generally obtained from the extracted packet header, and the flow identifier may be a quintuple, that is, composed of a source IP address, a destination IP address, a source port number, a destination port number, and a protocol type.
- the packet identifier is also generally obtained from the extracted packet header.
- the identifier field (Identification) field and the fragment offset (Fragment-offset) of the IP packet header can be determined as unique identifiers. Packet ID of the packet.
- Step 200 The measurement controller configures the measurement point.
- the configuration includes: the quintuple of the measured stream, the measurement start time, and the measurement end time. If you need to use sampling, you also need to provide a sampling algorithm.
- Step 205 The measurement points A and B are configured according to the configuration, and when the measurement start time is reached, the target message is extracted according to the quintuple of the measured flow and the timestamp information is attached; and the flow identifier and the message identifier are generated according to the content of the extracted message.
- Step 210 The measurement points A and B combine the flow identifier, the packet identifier, and the timestamp information into a packet summary data transmission measurement data collector.
- Step 215 The collector receives the packet digest data of the measurement points A and B, compares the packet digest data received from the measurement points VIII and B, matches the flow identifier and the packet identifier, compares the timestamps, and obtains a single packet. Delay. If a message identifier appears in the message digest data collected from measurement point A and does not appear in the message digest data collected from measurement point B, then the message is judged to be lost.
- Step 220 The collector calculates the network performance indicators in a calculation interval according to the delay of the single packet and the packet loss, that is, the average delay, the maximum delay, the minimum delay, the delay jitter, and the packet loss rate. Network unavailability, etc.
- Step 225 The collector sends performance indicators such as average delay, maximum delay, minimum delay, delay jitter, packet loss rate, and network unavailability to the measurement controller.
- Step 230 The measurement controller reports the performance index of the text network to the user or other operators. After the measurement end time is reached, the measurement point stops extracting the target message.
- the stream identifier is generated according to the extracted packet header information, when the measured stream is a data stream between the two networks, the stream identifier cannot be generated by using the extracted packet header information. For example: From 100.1.1.0/24 to 200.1.1.0/24, the description of this aggregated flow should be composed of "source network address 100.1.1.0/24, destination network address 200.1.1.0/24, priority information," The packet header does not contain mask information, so the stream identifier of this aggregated stream cannot be obtained from the packet header.
- the identifier is used as the text.
- the Identification field and the Fragment-offset can be used as fields that uniquely identify the message.
- both the identifier field and the fragment offset (Fragment-offset) are likely to be duplicated.
- the identifier field and the fragment offset cannot be used as the text identifier, and the generation of the packet identifier does not take into account the IPv6 environment, because there is no IPv4 identifier in the IPv6 header.
- the embodiment of the invention provides a network performance measurement method and system, which can solve the problem that the network performance measurement cannot be performed for the aggregated flow in the existing measurement; further solve the problem that the packet identification may not be obtained and the measurement is affected.
- a network performance measurement method includes the following steps:
- the measurement point extracts the packet belonging to the measured flow according to the flow description, and generates the packet summary data including the flow identifier and the packet identifier, and generates the packet summary data;
- the matching stream identifier and the ⁇ ⁇ identifier identify the scorpion digest data reported by different measurement points and belong to the same essay, and calculate the network performance metric according to the packet digest data.
- a network performance measurement method includes the following steps:
- the measurement point extracts the packet that belongs to the measured flow according to the flow description, generates a flow identifier according to the flow description, and generates message summary data including the flow identifier and the packet identifier, and reports the data;
- the matching flow identifier and the packet identifier determine the summary data of the packets that are reported by the different measurement points and belong to the same packet, and calculate the network performance indicator according to the packet summary data.
- a network performance measurement system includes: a measurement controller, a collector, and a measurement device; the measurement controller allocates a flow identifier for the measured flow according to the network performance measurement request, and configures the flow identifier and the flow description of the measured flow to Measuring device
- the measuring device extracts the packet belonging to the measured flow according to the flow description, generates the packet summary data including the flow identifier and the message identifier, and reports the packet summary data to the collector;
- the collector calculates a network performance indicator according to the reported message summary data, and sends the network performance indicator to the measurement controller.
- a network performance measurement system includes: a measurement controller, a collector, and a measurement device; the measurement controller configures a flow description of the flow to be measured according to a network performance measurement request;
- the measuring device extracts a packet belonging to the measured flow according to the flow description, generates a flow identifier according to the flow description, and generates message summary data including the flow identifier and the packet identifier, and reports the packet summary data to the collector;
- the network performance indicator is calculated according to the reported message summary data and sent to the measurement controller.
- the flow identifier is allocated by the measurement controller, or the flow point identifier is generated by the measurement point according to the flow description of the measured flow, instead of being obtained from the extracted target message, and therefore, the measurement of the micro flow and the aggregate flow can be supported.
- the content of the header of the extracted IP packet and the content of the payload part It can not only support IPv4 and IPv6 formats, but also support streaming (determination of a stream by IPv4 quintuple or IPv6 triple) and measurement of aggregated flows.
- FIG. 1 is a schematic structural view of a measurement system in a prior art
- FIG. 3 is a flow chart of implementing passive measurement by a process controller by assigning a process identifier according to an embodiment of the present invention
- FIG. 4 is a flow chart of passive measurement performed by a measurement point according to a flow description to generate a flow identifier according to an embodiment of the present invention. detailed description
- the measurement controller assigns a unique flow identifier to the flow description of each measured flow according to the measured request, '
- the flow description and the flow identifier are configured to the measurement point, and the measurement point extracts the packet belonging to the measured flow according to the flow description, attaches the flow identifier to the generated message summary data, and reports it to the collector; the collector matches the flow identifier and
- the packet identifier identifies the packet digest data that is reported by the different measurement points and belongs to the same packet, and calculates the network performance indicator according to the packet digest data and reports it to the measurement controller.
- the measurement controller and the acquirer can be separate entities or the same entity.
- the description of the measured stream sent by the measurement controller to the measurement point can be one of the following descriptions:
- Source IP address ⁇ IP address, destination IP address, source port number, destination port number, protocol type (applies only to IPv4).
- MPLS Multi-Protocol Label Switching
- the IPv4 packet or the IPv6 packet has different contents in the header of the packet and the payload of the packet in different packets. Therefore, these differences can be compared. Large fields are grouped together to form a suffix identifier that uniquely identifies a suffix.
- identifier and the Fragment-offset For IPv4, you can use the identifier and the Fragment-offset to combine the part of the protocol, the source IP address, and the destination IP address. Or the contents of all the fields to generate a message identifier; further, a part of the payload data (generally, the first 20 bytes of the payload portion) may be added to generate a message identifier.
- the Payload Length, the Next Header, the Source IP address, and the Destination IP address are used to generate the 4 ⁇ identifier; further, it can be added
- the upper partial payload data (generally, the first 20 bytes of the payload portion can be generated) generates a ⁇ text identifier.
- a suitable function can be used (ie, it can avoid the occurrence of message identification conflicts, that is, the message identification is unique during the measurement), and generate a text identifier according to the above fields, for example, using a loop.
- Check code function CRC32
- compression function HASH
- hash function HASH
- the process of implementing the passive measurement by the flow controller by the measurement controller is as follows: 'Step 300, the measurement controller assigns a flow identifier to the measured flow and configures the measurement point.
- the configuration includes: a flow identifier of the measured flow, a flow description of the measured flow, a measurement start time, and a measurement end time. If you need to apply sampling, you also need to provide a sampling algorithm.
- the flow is described as one of the foregoing descriptions.
- Step 305 When the measurement start time is reached, the measurement points A and B respectively extract the target message according to the measured flow description and attach the time stamp information, and at the same time, adopt an identifier (Identification), a fragment offset (Fragment-offset), The protocol type, the source IP address, and the destination IP address are used to generate the suffix identifier (using IPv4 as an example).
- Step 310 The measurement points A and B respectively combine the flow identifier, the packet identifier, and the timestamp information into the message summary data and send the data to the collector.
- Step 315 The collector receives the packet digest data of the measurement points A and B, compares the packet digest data received from the measurement points A and B, matches the flow identifier and the packet identifier, compares the time stamps, and obtains a single packet. Delay.
- Step 320 The collector calculates the network performance indicators in a calculation interval according to the delay of the single packet and the packet loss, that is, the average delay, the maximum delay, the minimum delay, the delay jitter, and the packet loss rate. Network unavailability, etc.
- Step 325 The collector sends performance indicators such as average delay, maximum delay, minimum delay, delay jitter, packet loss rate, and network unavailability to the measurement controller.
- Step 330 The measurement controller reports the performance index of the text network to the user or other operators. After the measurement end time is reached, the measurement point stops extracting the target message.
- the flow identifier is obtained by the flow description of the measured flow that is sent by the measurement controller, instead of from the extracted target information. obtain.
- turbulence measurements can be supported while also supporting the measurement of aggregated flows.
- the field described by the measured stream can be subjected to a function operation to calculate a stream identifier.
- This calculation function can be a cyclic redundancy code (CRC16, CRC32) function, a compression function or a hash function (HASH), as long as the stream identifier generated by the function is guaranteed to be unique during the measurement period. In this way, a fixed-length stream identifier can be generated, which also helps to match the packet summary data.
- Step 400 The measurement controller configures the measurement point.
- the configuration includes: a flow description of the measured flow, a measurement start time, and a measurement end time. If you need to apply sampling, you also need to provide a sampling algorithm.
- the flow is described as one of the foregoing descriptions.
- Step 405 When the measurement start time is reached, the measurement points A and B respectively extract the target message according to the measured flow description and attach the time stamp information; meanwhile, generate a flow identifier by using a cyclic redundancy code function according to the field in the measured description, and Use the Payload Length, Next Header, Source IP address, Destination IP address, and the first 20 bytes of the payload to generate the text identifier (in IPv6). example).
- Step 410 The measurement points A and B respectively combine the flow identifier, the packet identifier, and the timestamp information into the packet summary data and send the data to the collector.
- Step 415 The collector receives the packet summary data of the measurement points A and B, compares the packet summary data received from the measurement points A and B, matches the flow identifier and the identifier, compares the time stamp, and obtains a single ⁇ The delay of the text.
- Step 420 The collector calculates the network performance indicators in a calculation interval according to the delay of the single packet and the packet loss, that is, the average delay, the maximum delay, the minimum delay, the delay jitter, and the packet loss rate. And network unavailability, etc.
- Step 425 The collector sends performance indicators such as average delay, maximum delay, minimum delay, delay jitter, packet loss rate, and network unavailability to the measurement controller.
- Step 430 The measurement controller reports the performance indicators of the network to the user or other operators. After the measurement end time is reached, the measurement point stops extracting the target message.
- IPv4 and IPv6 formats can be used to support not only IPv4 and IPv6 formats, but also microflow (determining a stream by IPv4, quintuple or IPv6 or triple) and aggregated flow measurements.
- the flow identifier is allocated by the measurement controller, or the flow point identifier is generated by the measurement point according to the flow description of the measured flow, instead of being obtained from the extracted target message, and therefore, the micro flow can be supported.
- the packet identifier is generated according to the content in the header of the extracted IP packet and the content of the payload portion, and not only supports the IPv4 and IPv6 formats, but also supports the microflow (determining a flow by the IPv4 quintuple or the IPv6 triplet) And the measurement of the aggregate flow.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
A method for network performance measurement, includes the steps of: distributing the flow identification for the measured flow according to the network performance measurement request, configuring the flow identification and the description of the measured flow to the measurement point; the measurement point extracting the packet of the measured flow according to the description of measured flow, generating packet abstract data including the flow identification and the packet identification and reporting it; and calculating the network performance index according to the packet abstract data. The present invention also includes a system for network performance measurement.
Description
网络性能测量方法及系统 Network performance measurement method and system
技术领域 Technical field
本发明涉及通信领域的测量技术, 尤其涉及网络性能测量方法及系统。 背景技术 The present invention relates to measurement techniques in the field of communications, and more particularly to methods and systems for measuring network performance. Background technique
随着 Internet技术和网络业务的飞速发展, 用户对网络资源的需求空前增 长, 网络也变得越来越复杂。 不断增加的网络用户和应用, 导致网络负担沉 重, 网络设备超负荷运转, 从而引起网络性能下降。 这就需要对网络的性能 指标进行提取与分析, 对网络性能进行改善和提高, 因此网络性能测量便应 运而生。 通过测量能够发现网络瓶颈, 优化网络配置, 并进一步发现网络中 可能存在的潜在危险, 更加有效地进行网络性能管理, 提供网络服务质量的 验证和控制。 对服务提供商的服务质量指标进行量化、 比较和验证, 是网络 性能测量的主要目的。 With the rapid development of Internet technologies and network services, users' demand for network resources has grown at an unprecedented rate, and the network has become more and more complex. Increasing network users and applications result in heavy network burdens and overloaded network devices, causing network performance degradation. This requires the extraction and analysis of network performance indicators to improve and improve network performance, so network performance measurement has emerged. Through measurement, network bottlenecks can be discovered, network configuration can be optimized, and potential risks in the network can be further discovered, network performance management can be performed more effectively, and network service quality verification and control can be provided. Quantifying, comparing, and verifying service provider quality of service metrics is the primary goal of network performance measurement.
最常见的 IP网络性能测量方法有两类: 主动测量和被动测量。 主动测量 是在选定的测量点上利用测量工具有目的地主动产生测量流量, 注入网络, 并根据测量数据流的传送情况来分析网絡的性能。 被动测量是指在链路或设 备(如路由器, 交换机等) 上利用测量设备对网络进行监测, 而不需要产生 多余流量的测量方法。 There are two main types of IP network performance measurement methods: active measurement and passive measurement. Active measurement is the use of measurement tools to actively generate measurement traffic at selected measurement points, inject into the network, and analyze the performance of the network based on the measurement data flow. Passive measurement refers to the measurement of the network using measurement equipment on links or devices (such as routers, switches, etc.) without the need to generate excess traffic.
网络性能测量从方向上来划分, 还可以分为单向网络性能测量和往返网 络性能测量。 单向网络性能测量就是测量从一个测量点 A到另一个测量点 B 一个方向的网络性能指标。 往返网络性能测量是测量从一个测量点 A到另一 个测量点 B再返回测量点 A的网络性能指标。 对于两点之间的通信, 在往返 两个方向上可能经过不同的路径, 或者同一个路径往返两个方向的 QoS是不 相同的, 因此单向网络性能并不等于筒单的将往返时延除以二。 在本发明如 非特別指出, 网络性能测量均为单向网络性能测量。 Network performance measurements are divided into directions and can be divided into one-way network performance measurement and round-trip network performance measurement. Unidirectional network performance measurement is a measure of network performance in one direction from one measurement point A to another. The round-trip network performance measurement measures network performance indicators from one measurement point A to another measurement point B and back to measurement point A. For the communication between two points, the QoS may be different in both directions of the round-trip, or the QoS of the same path to and from the two directions is different, so the one-way network performance is not equal to the round-trip delay of the single ticket. Divide by two. Unless otherwise specified in the present invention, network performance measurements are all one-way network performance measurements.
如图 1所示,被动测量系统一般包括测量控制器、采集器和测量装置(或 称测量点)构成。 测量控制器的功能如下:
1、 对测量点进行配置, 包括: 被测流的五元组, 测量持续的时间。 如果 使用采样, 还需配置测量点的采样算法。 As shown in FIG. 1, the passive measurement system generally includes a measurement controller, a collector, and a measurement device (or a measurement point). The function of the measurement controller is as follows: 1. Configure the measurement points, including: The quintuple of the measured flow, the duration of the measurement. If you use sampling, you also need to configure the sampling algorithm for the measurement points.
■ 2、 接收采集器上报的网絡性能指标, 以及向用户或者其他运营商网络的 测量控制器报文本网络的性能指标。 ■ 2. Receive the network performance indicators reported by the collector and the performance indicators of the text network to the measurement controller of the user or other carrier network.
3、 向其它运营商网络的测量控制器发送测量倩求。 3. Send measurement requests to the measurement controllers of other carrier networks.
测量装置的功能如下: The function of the measuring device is as follows:
1、 根据测量控制器的配置提取所有或者部分属于被测流的报文(如果使 用采样的方法减少提取报文的数量, 那么测量点仅根据采样算法提取被测流 的一部分报文)。 1. Extract all or part of the packets belonging to the measured stream according to the configuration of the measurement controller (if the sampling method is used to reduce the number of extracted packets, the measurement points only extract a part of the packets of the measured stream according to the sampling algorithm).
2、 时钟同步。 进行单向性能测量, 要求测量点之间必须时钟同步。 2. Clock synchronization. Performing a one-way performance measurement requires that the measurement points must be clock synchronized.
' 3、 如果使用采样, 根据采用算法提取被测流的一部分报文。 ' 3. If sampling is used, a part of the message of the measured stream is extracted according to the algorithm.
4、 在提取的目标报文中附加一个时间戳信息。 4. Attach a timestamp information to the extracted target message.
5、 根据提取的报文生成报文摘要数据, 在报文摘要数据中至少要包括: -流标识, 用来唯一标识被测流。 不同测量点的报文摘要数据汇聚到采集 器后, 根据流标识来匹配从不同测量点提取的报文摘要数椐, 确定不同测量 点采集的同一个被测流。 5. Generate message digest data according to the extracted packet, and at least the message digest data includes: - a stream identifier, which is used to uniquely identify the stream to be measured. After the packet digest data of different measurement points is aggregated to the collector, the number of packet digests extracted from different measurement points is matched according to the flow identifier to determine the same measured stream collected by different measurement points.
-报文标识, 用来唯一标识属于同一个被测流的 艮文。 在 ^艮文摘要数据汇 聚到采集后, 首先根据流标识匹配从不同测量点提取的报文摘要数据, 再根 ' 据 4艮文标识进行匹配, 从而确定不同测量点所采集的同一个目标¾_文。 - Message identification, which is used to uniquely identify the essays belonging to the same measured stream. After the summary data of the 艮 汇 汇 汇 汇 汇 汇 , 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要 摘要_ text.
-时间戳, 用来测量点时延。 根据不同测量点的报文摘要数据中的时间戳 信息, 确定两个测量点之间时延。 - Timestamp, used to measure point delay. The delay between two measurement points is determined according to the timestamp information in the message summary data of different measurement points.
采集器的功能如下: The functions of the collector are as follows:
1、 存储从测量点发送的报文摘要数据。 1. Store the message summary data sent from the measurement point.
2、 根据不同测量点的报文摘要数据, 计算每一个报文的时延以及是否丟 失 (如果上游测量点提取到报文, 在下游测量点没有提取到, 那么就判断这 个报文在传送过程中丢失了)。 2. Calculate the delay of each packet and whether it is lost according to the packet summary data of different measurement points. (If the upstream measurement point extracts the message and does not extract it at the downstream measurement point, then it is judged that the message is in the transmission process. Lost in it).
' 3、 根据每一个^ ^艮文的时延和丟失与否, 可以计算出一个时间段内 (这个
时间段称作计算间隔 Evaluation lnterval )的平均时延、 最大时延、 最小时延、 时延抖动、 丢包率以及网络不可用性等网络性能指标。 ' 3, according to the delay and loss of each ^ ^ 艮 text, can be calculated within a period of time (this The time period is called the calculation interval (Evaluation lnterval), and the network performance indicators such as average delay, maximum delay, delay, delay jitter, packet loss rate, and network unavailability.
4、 将性能指标报告给测量控制器。 4. Report the performance indicators to the measurement controller.
在现有技术中, 流标识一般从提取的报文头中获得, 流标识可以是五元 组, 即由源 IP地址、 目的 IP地址、 源端口号、 目的端口号和协议类型组成。 报文标识一般也从提取的报文头中获得; 对于 IPv4报文,可以将 IP报文头部 的标识符(Identification )字段和分片偏移量(Fragment-offset ) 故为唯一标识 被测报文的报文标识。 In the prior art, the flow identifier is generally obtained from the extracted packet header, and the flow identifier may be a quintuple, that is, composed of a source IP address, a destination IP address, a source port number, a destination port number, and a protocol type. The packet identifier is also generally obtained from the extracted packet header. For the IPv4 packet, the identifier field (Identification) field and the fragment offset (Fragment-offset) of the IP packet header can be determined as unique identifiers. Packet ID of the packet.
如图 2所示, 在现有技术中, 针对一个测量任务其被动测量的具体流程 如下: As shown in FIG. 2, in the prior art, the specific process of passive measurement for a measurement task is as follows:
步骤 200、 测量控制器配置测量点。 配置包括: 被测流的五元组、 测量开 始时间, 测量结束时间。 如果需要症用采样, 还需提供采样算法。 Step 200: The measurement controller configures the measurement point. The configuration includes: the quintuple of the measured stream, the measurement start time, and the measurement end time. If you need to use sampling, you also need to provide a sampling algorithm.
步骤 205、 测量点 A、 B根据配置, 到达测量开始时间时, 根据被测流的 五元组提取目标报文并附着时间戳信息; 并根据提取报文的内容生成流标识 和报文标识。 Step 205: The measurement points A and B are configured according to the configuration, and when the measurement start time is reached, the target message is extracted according to the quintuple of the measured flow and the timestamp information is attached; and the flow identifier and the message identifier are generated according to the content of the extracted message.
步骤 210、 测量点 A、 B将流标识、 报文标识和时间戳信息合并为报文摘 要数据发送测量数据采集器。 Step 210: The measurement points A and B combine the flow identifier, the packet identifier, and the timestamp information into a packet summary data transmission measurement data collector.
步驟 215、 采集器接受测量点 A、 B的报文摘要数据, 比较从测量点八、 B 收到的报文摘要数据, 匹配流标识和报文标识, 比较时间戳, 从而获得针 对单个报文的时延。 如果一个报文标识出现在从测量点 A收集的报文摘要数 据中而没有出现在从测量点 B收集的报文摘要数据中, 那么, 该报文即判定 为丢失。 Step 215: The collector receives the packet digest data of the measurement points A and B, compares the packet digest data received from the measurement points VIII and B, matches the flow identifier and the packet identifier, compares the timestamps, and obtains a single packet. Delay. If a message identifier appears in the message digest data collected from measurement point A and does not appear in the message digest data collected from measurement point B, then the message is judged to be lost.
步骤 220、 采集器根据单个报文的时延、 和报文丟失情况, 计算在一个计 算间隔的网络性能指标, 即平均时延、 最大时延、 最小时延、 时延抖动、 丢 包率以及网络不可用性等。 Step 220: The collector calculates the network performance indicators in a calculation interval according to the delay of the single packet and the packet loss, that is, the average delay, the maximum delay, the minimum delay, the delay jitter, and the packet loss rate. Network unavailability, etc.
步骤 225、 采集器向测量控制器发送平均时延、 最大时延、 最小时延、 时 延抖动、 丢包率以及网络不可用性等性能指标。
步驟 230、 测量控制器向用户或者其他运营商报文本网络的性能指标。 到达测量结束时间后, 测量点停止提取目标报文。 Step 225: The collector sends performance indicators such as average delay, maximum delay, minimum delay, delay jitter, packet loss rate, and network unavailability to the measurement controller. Step 230: The measurement controller reports the performance index of the text network to the user or other operators. After the measurement end time is reached, the measurement point stops extracting the target message.
虽然上述方案能够进行被动测量, 但发明人在研发过程发现以下问题: Although the above scheme is capable of passive measurement, the inventors found the following problems during the development process:
( 1 )现有的被动测量没有考虑对聚合流的测量。 虽然五元组可以用来描 述两个主机之间的数据流, 但被测流可能是两个网络之间的数据流或者其他 形式的聚合流, 此时就无法采用五元组来描述被测流, 因此无法进行测量。 (1) Existing passive measurements do not take into account the measurement of the aggregate flow. Although a quintuple can be used to describe the data flow between two hosts, the measured flow may be a data flow between two networks or other forms of aggregated flow. At this point, the quintuple cannot be used to describe the measured Flow, so measurement is not possible.
( 2 ) 由于根据提取的报文头部信息生成流标识, 因此, 当被测流是两个 网络之间的数据流时, 则无法用提取到的报文头部信息生成流标识。 例如: 从 100.1.1.0/24到 200.1.1.0/24,对于这个聚合流的描述应该是由 "源网络地址 100.1.1.0/24, 目的网络地址 200.1.1.0/24, 优先级信息,, 组成的元组。 报文头 部不包含掩码信息, 因此无法从报文头部获得这个聚合流的流标识。 (2) Since the stream identifier is generated according to the extracted packet header information, when the measured stream is a data stream between the two networks, the stream identifier cannot be generated by using the extracted packet header information. For example: From 100.1.1.0/24 to 200.1.1.0/24, the description of this aggregated flow should be composed of "source network address 100.1.1.0/24, destination network address 200.1.1.0/24, priority information," The packet header does not contain mask information, so the stream identifier of this aggregated stream cannot be obtained from the packet header.
( 3 )根据 IPv4 报文头中标识符 (Identification ) 字段和分片偏移量 ( Fragment-offset ) 作为 艮文标识。 对于两个主机之间的通信, 标识符 ( Identification )字段和分片偏移量(Fragment-offset )可以作为唯一标识报文 的字段。 但是对于一个聚合流, 可能存在多个主机之间的通信, 那么标识符 ( Identification )字段和分片偏移量(Fragment-offset )都是存在重复的可能性 的。 在这种情况下, 标识符字段和分片偏移量不能作为 艮文标识, 而且报文 标识的生成也没有考虑到 IPv6的环境, 因为在 IPv6的报文头中已经没有了 IPv4的标识符字段和分片偏移量字段。 发明内容 (3) According to the identifier (Identification) field and the fragment offset (Fragment-offset) in the IPv4 header, the identifier is used as the text. For communication between two hosts, the Identification field and the Fragment-offset can be used as fields that uniquely identify the message. However, for an aggregated stream, there may be communication between multiple hosts, so both the identifier field and the fragment offset (Fragment-offset) are likely to be duplicated. In this case, the identifier field and the fragment offset cannot be used as the text identifier, and the generation of the packet identifier does not take into account the IPv6 environment, because there is no IPv4 identifier in the IPv6 header. Field and fragment offset fields. Summary of the invention
本发明实施例提供一种网絡性能测量方法及系统, 以解决现有被测量中 无法针对聚合流进行网络性能测量的问题; 进一步解决可能无法得到报文标 识而影响测量的问题。 The embodiment of the invention provides a network performance measurement method and system, which can solve the problem that the network performance measurement cannot be performed for the aggregated flow in the existing measurement; further solve the problem that the packet identification may not be obtained and the measurement is affected.
一种网络性能测量方法, 包括如下步驟: A network performance measurement method includes the following steps:
>据网络性能测量请求为被测流分配流标识, 并将流标识和被测流的流 描述配置到测量点;
测量点根据流描述提取属于被测流的报文, 生成包含流标识和报文标识 的报文摘要数据并上 4艮; > assigning a flow identifier to the measured flow according to the network performance measurement request, and configuring the flow identifier and the flow description of the measured flow to the measurement point; The measurement point extracts the packet belonging to the measured flow according to the flow description, and generates the packet summary data including the flow identifier and the packet identifier, and generates the packet summary data;
匹配流标识和 ^艮文标识确定不同测量点上报的且属于同一艮文的 艮文摘 要数据, 并依据报文摘要数据计算网络性能指标。 The matching stream identifier and the 艮 标识 identifier identify the scorpion digest data reported by different measurement points and belong to the same essay, and calculate the network performance metric according to the packet digest data.
一种网络性能测量方法, 包括如下步骤: A network performance measurement method includes the following steps:
根据网络性能测量将被测流的流描述配置到测量点; Configuring a flow description of the measured flow to the measurement point according to the network performance measurement;
测量点根据流描述提取属于被测流的报文, 根据所述流描述生成流标识, 以及生成包含该流标识和报文标识的报文摘要数据并上报; The measurement point extracts the packet that belongs to the measured flow according to the flow description, generates a flow identifier according to the flow description, and generates message summary data including the flow identifier and the packet identifier, and reports the data;
匹配流标识和报文标识确定不同测量点上报的且属于同一报文的报文摘 要数据, 并依据报文摘要数据计算网络性能指标。 The matching flow identifier and the packet identifier determine the summary data of the packets that are reported by the different measurement points and belong to the same packet, and calculate the network performance indicator according to the packet summary data.
一种网络性能测量系统, 包括: 测量控制器、 采集器和测量装置; 所述测量控制器根据网络性能测量请求为被测流分配流标识, 并将流标 识和被测流的流描述配置到测量装置; A network performance measurement system includes: a measurement controller, a collector, and a measurement device; the measurement controller allocates a flow identifier for the measured flow according to the network performance measurement request, and configures the flow identifier and the flow description of the measured flow to Measuring device
所述测量装置根据流描述提取属于被测流的报文, 生成包含流标识和报 文标识的报文摘要数据并上报所述采集器; The measuring device extracts the packet belonging to the measured flow according to the flow description, generates the packet summary data including the flow identifier and the message identifier, and reports the packet summary data to the collector;
所述采集器根据上报的报文摘要数据计算网絡性能指标, 并发送给测量 控制器。 The collector calculates a network performance indicator according to the reported message summary data, and sends the network performance indicator to the measurement controller.
一种网络性能测量系统, 包括: 测量控制器、 采集器和测量装置; 所述测量控制器根据网络性能测量请求将被测流的流描述配置到测量 ; A network performance measurement system includes: a measurement controller, a collector, and a measurement device; the measurement controller configures a flow description of the flow to be measured according to a network performance measurement request;
所述测量装置根据流描述提取属于被测流的报文和根据所述流描述生成 流标识, 以及生成包含流标识和报文标识的报文摘要数据并上报所述采集器; 所述采集器根据上报的报文摘要数据计算网络性能指标, 并发送给测量 控制器。 The measuring device extracts a packet belonging to the measured flow according to the flow description, generates a flow identifier according to the flow description, and generates message summary data including the flow identifier and the packet identifier, and reports the packet summary data to the collector; The network performance indicator is calculated according to the reported message summary data and sent to the measurement controller.
本发明实施例中由测量控制器分配流标识, 或者由测量点根据被测流的 流描述生成流标识, 而不是从提取的目标报文中获得, 因此, 能够支持微流 和聚合流的测量。 根据提取的 IP报文的报文头中的内容和净荷部分的内容生
成^艮文标识, 不仅可以支持 IPv4和 IPv6格式, 同时支持 流(由 IPv4五元 组或者 IPv6三元组确定一个流)和聚合流的测量。 附图说明 In the embodiment of the present invention, the flow identifier is allocated by the measurement controller, or the flow point identifier is generated by the measurement point according to the flow description of the measured flow, instead of being obtained from the extracted target message, and therefore, the measurement of the micro flow and the aggregate flow can be supported. . According to the content of the header of the extracted IP packet and the content of the payload part It can not only support IPv4 and IPv6 formats, but also support streaming (determination of a stream by IPv4 quintuple or IPv6 triple) and measurement of aggregated flows. DRAWINGS
图 1为现有扶术中测量系统的结构示意图; 1 is a schematic structural view of a measurement system in a prior art;
图 2为现有技术中实现被动测量的流程图; 2 is a flow chart of implementing passive measurement in the prior art;
, 图 3 为本发明实施例中由测量控制器分配流程标识实现被动测量的流程 图; FIG. 3 is a flow chart of implementing passive measurement by a process controller by assigning a process identifier according to an embodiment of the present invention;
图 4 为本发明实施例中测量点根据流描述生成流标识实现被动测量的流 程图。 具体实施方式 FIG. 4 is a flow chart of passive measurement performed by a measurement point according to a flow description to generate a flow identifier according to an embodiment of the present invention. detailed description
实施例一 Embodiment 1
参阅图 1 所示, 在测量控制系统中, 测量控制器除了完成现有的基本功 能外, 还才艮据被测量请求为每一个被测流的流描述分配一个唯一的流标识, ' 并将流描述和流标识配置到测量点, 测量点则按流描述提取属于被测流的报 文, 将流标识附加在生成的报文摘要数据中并上报到釆集器; 采集器匹配流 标识和报文标识确定不同测量点上报的且属于同一报文的报文摘要数据, 并 依据报文摘要数据计算网络性能指标并上报到测量控制器。 测量控制器与采 集器可以是相互独立的实体, 也可以是同一个实体。 这样, 只要在测量控制 器和测量点保存分配的流标识和流描述的对应表即可, 采集器进行性能指标 的计算根据流标识的具体含义无关, 只要能唯一标识一个被测流即可, 从而 可以筒化实现和提高效率。 Referring to FIG. 1, in the measurement control system, in addition to completing the existing basic functions, the measurement controller assigns a unique flow identifier to the flow description of each measured flow according to the measured request, ' The flow description and the flow identifier are configured to the measurement point, and the measurement point extracts the packet belonging to the measured flow according to the flow description, attaches the flow identifier to the generated message summary data, and reports it to the collector; the collector matches the flow identifier and The packet identifier identifies the packet digest data that is reported by the different measurement points and belongs to the same packet, and calculates the network performance indicator according to the packet digest data and reports it to the measurement controller. The measurement controller and the acquirer can be separate entities or the same entity. In this way, as long as the corresponding table of the flow identifier and the flow description is saved in the measurement controller and the measurement point, the calculation of the performance index of the collector has nothing to do according to the specific meaning of the flow identifier, as long as it can uniquely identify a measured flow, This makes it possible to achieve and improve efficiency.
' 为了支持不同粒度流的性能测量(包括微流和聚合流), 以及同时支持' To support performance measurement of different granularity flows (including microflow and aggregated flows), and support at the same time
IPv4和 IPv6,测量控制器下发给测量点的被测流的描述可以是下列描述之一:For IPv4 and IPv6, the description of the measured stream sent by the measurement controller to the measurement point can be one of the following descriptions:
( 1 )源 IP地址, 目的 IP地址, 源端口号, 目的端口号, 协议类型 (仅 应用 IPv4 )。 (1) Source IP address, destination IP address, source port number, destination port number, protocol type (applies only to IPv4).
( 2 ) 源网络地址, 目的网络地址, 差分编码点 (DiffServ Code Point,
DSCP ) (仅应用 IPv4 )。 (2) source network address, destination network address, differential code point (DiffServ Code Point, DSCP) (applies only to IPv4).
( 3 ) 源 IP地址, 目的 IP地址, 流标签(仅应用 IPv6 )0 (3) Source IP address, destination IP address, flow label (applies only to IPv6) 0
( 4 ) 源网络地址, 目的网络地址, 流量类型 (Traffic Class ) (仅应用 IPv6 λ (4) Source network address, destination network address, traffic type (Traffic Class) (applies only to IPv6 λ
( 5 ) 多协议标签交换 ( MPLS )标签。 (5) Multi-Protocol Label Switching (MPLS) label.
( 6 )其他聚合流的描述。 (6) Description of other aggregated flows.
根据 IP报文格式可以知道, 无论是 IPv4报文还是 IPv6报文, 对于不同 的报文其报文头部和报文净荷部分某些字段的内容区别比较大, 因此, 可以 将这些区别比较大的字段放在一起组合成为一个艮文标识, 用来唯一标识一 个 4艮文。 According to the IP packet format, it can be known that the IPv4 packet or the IPv6 packet has different contents in the header of the packet and the payload of the packet in different packets. Therefore, these differences can be compared. Large fields are grouped together to form a suffix identifier that uniquely identifies a suffix.
对于 IPv4 来说, 可以用标识符 ( Identification ) 和分片偏移量 ( Fragment- offset )结合十办议类型 ( Protocol ), 源地址 ( Source IP address ), 目的地址(Destination IP address ) 中的部分或全部字段的内容来生成报文标 识; 进一步的, 可以再加上部分净荷数据(一般地, 净荷部分的前 20个字节 即可)生成艮文标识、。 For IPv4, you can use the identifier and the Fragment-offset to combine the part of the protocol, the source IP address, and the destination IP address. Or the contents of all the fields to generate a message identifier; further, a part of the payload data (generally, the first 20 bytes of the payload portion) may be added to generate a message identifier.
对于 IPv6来说, 净荷长度 ( Payload Length ), 下一个头部( Next Header ), 源地址( Source IP address ), 目的地址( Destination IP address )来生成 4艮文标 识; 进一步的, 可以再加上部分净荷数据(一般地, 净荷部分的前 20个字节 即可)生成 ^^文标识。 For IPv6, the Payload Length, the Next Header, the Source IP address, and the Destination IP address are used to generate the 4 艮 identifier; further, it can be added The upper partial payload data (generally, the first 20 bytes of the payload portion can be generated) generates a ^^ text identifier.
为了尽可能节约存储空间和网络资源, 可以釆用合适的函数(即能够避 免产生报文标识冲突, 也就是在测量期间该报文标识唯一)根据上述字段生 成一个艮文标识, 如, 采用循环校验码函数(CRC32 )、 压缩函数或者哈希函 数(HASH )等。 In order to save storage space and network resources as much as possible, a suitable function can be used (ie, it can avoid the occurrence of message identification conflicts, that is, the message identification is unique during the measurement), and generate a text identifier according to the above fields, for example, using a loop. Check code function (CRC32), compression function or hash function (HASH).
参阅图 3所示, 由测量控制器分配流标识实现被动测量的过程如下: '步骤 300、 测量控制器为被测流分配流标识并配置测量点。 配置包括: 被 测流的流标识、 被测流的流描述、 测量开始时间, 测量结束时间。 如果需要 应用采样, 还需提供采样算法。 流描述为前述的描述之一。
步骤 305、 到达测量开始时间时, 测量点 A、 B分别根据被测流描述提取 目标报文并附着时间戳信息, 同时, 采用标识符(Identification )、 分片偏移量 ( Fragment-offset )、 寸办议类型 ( Protocol )、 源地址 ( Source IP address )和目 的地址( Destination IP address )生成艮文标识 (以 IPv4为例)。 Referring to FIG. 3, the process of implementing the passive measurement by the flow controller by the measurement controller is as follows: 'Step 300, the measurement controller assigns a flow identifier to the measured flow and configures the measurement point. The configuration includes: a flow identifier of the measured flow, a flow description of the measured flow, a measurement start time, and a measurement end time. If you need to apply sampling, you also need to provide a sampling algorithm. The flow is described as one of the foregoing descriptions. Step 305: When the measurement start time is reached, the measurement points A and B respectively extract the target message according to the measured flow description and attach the time stamp information, and at the same time, adopt an identifier (Identification), a fragment offset (Fragment-offset), The protocol type, the source IP address, and the destination IP address are used to generate the suffix identifier (using IPv4 as an example).
步骤 310、 测量点 A、 B分别将流标识、 报文标识和时间戳信息合并为报 文摘要数据发送给采集器。 Step 310: The measurement points A and B respectively combine the flow identifier, the packet identifier, and the timestamp information into the message summary data and send the data to the collector.
步驟 315、 采集器接受测量点 A、 B的报文摘要数据, 比较从测量点 A、 B 收到的报文摘要数据, 匹配流标识和报文标识, 比较时间戳, 从而获得针 对单个报文的时延。 Step 315: The collector receives the packet digest data of the measurement points A and B, compares the packet digest data received from the measurement points A and B, matches the flow identifier and the packet identifier, compares the time stamps, and obtains a single packet. Delay.
如果一个报文标识出现在从测量点 A收集的报文摘要数据中而没有出现 在从测量点 B收集的报文摘要数据中, 则判定该报文丟失。 If a message identifier appears in the message digest data collected from measurement point A and does not appear in the message digest data collected from measurement point B, it is determined that the message is lost.
步骤 320、 采集器根据单个报文的时延、 和报文丟失情况, 计算在一个计 算间隔的网络性能指标, 即平均时延、 最大时延、 最小时延、 时延抖动、 丟 包率以及网络不可用性等。 Step 320: The collector calculates the network performance indicators in a calculation interval according to the delay of the single packet and the packet loss, that is, the average delay, the maximum delay, the minimum delay, the delay jitter, and the packet loss rate. Network unavailability, etc.
步骤 325、 采集器向测量控制器发送平均时延、 最大时延、 最小时延、 时 延抖动、 丟包率以及网络不可用性等性能指标。 Step 325: The collector sends performance indicators such as average delay, maximum delay, minimum delay, delay jitter, packet loss rate, and network unavailability to the measurement controller.
步骤 330、 测量控制器向用户或者其他运营商报文本网络的性能指标。 在到达测量结束时间后, 测量点停止提取目标报文。 Step 330: The measurement controller reports the performance index of the text network to the user or other operators. After the measurement end time is reached, the measurement point stops extracting the target message.
实施例二 Embodiment 2
参阅图 1 所示, 本实施例的测量控制系统中, 流标识由测量点才 居测量 控制器下发的被测流的流描述来获得, 而不是从提取的目标 4艮文头部信息中 获得。 这样, 可以支持敖流的测量, 同时还可以支持聚合流的测量。 测量点 将可以将被测流描述的字段经过函数运算, 计算出一个流标识。 这个计算函 数可以是循环冗余码( CRC16, CRC32 )函数,压缩函数或者哈希函数( HASH ), 只要保证函数生成的流标识在测量期间唯一即可。 这样, 可以生成一个固定 长度的流标识, 也有利于匹配报文摘要数据。 As shown in FIG. 1 , in the measurement control system of the embodiment, the flow identifier is obtained by the flow description of the measured flow that is sent by the measurement controller, instead of from the extracted target information. obtain. In this way, turbulence measurements can be supported while also supporting the measurement of aggregated flows. Measuring point The field described by the measured stream can be subjected to a function operation to calculate a stream identifier. This calculation function can be a cyclic redundancy code (CRC16, CRC32) function, a compression function or a hash function (HASH), as long as the stream identifier generated by the function is guaranteed to be unique during the measurement period. In this way, a fixed-length stream identifier can be generated, which also helps to match the packet summary data.
被测流的流描述以及 文生成方式与实施例一中的方式相同。
参阅图 4所示, 测量点根据流描述生成流标识实现被动测量的过程如下: 步骤 400、 测量控制器配置测量点。 配置包括: 被测流的流描述、 测量开 始时间, 测量结束时间。 如果需要应用采样, 还需提供采样算法。 流描述为 前述的描述之一。 The flow description of the measured stream and the manner of generating the text are the same as those in the first embodiment. Referring to FIG. 4, the process of generating a flow identifier according to the flow description to implement passive measurement is as follows: Step 400: The measurement controller configures the measurement point. The configuration includes: a flow description of the measured flow, a measurement start time, and a measurement end time. If you need to apply sampling, you also need to provide a sampling algorithm. The flow is described as one of the foregoing descriptions.
步骤 405、 到达测量开始时间时, 测量点 A、 B分别根据被测流描述提取 目标报文并附着时间戳信息; 同时, 根据被测描述中的字段采用循环冗余码 函数生成流标识, 以及采用净荷长度 ( Payload Length )、 下一个头部 (Next Header )、 源地址 ( Source IP address ), 目的地址 ( Destination IP address )和 净荷部分的前 20个字节生成 文标识(以 IPv6为例 )。 Step 405: When the measurement start time is reached, the measurement points A and B respectively extract the target message according to the measured flow description and attach the time stamp information; meanwhile, generate a flow identifier by using a cyclic redundancy code function according to the field in the measured description, and Use the Payload Length, Next Header, Source IP address, Destination IP address, and the first 20 bytes of the payload to generate the text identifier (in IPv6). example).
步骤 410、 测量点 A、 B分别将流标识、 报文标识和时间戳信息合并为报 文摘要数据发送给采集器。 Step 410: The measurement points A and B respectively combine the flow identifier, the packet identifier, and the timestamp information into the packet summary data and send the data to the collector.
步骤 415、 采集器接受测量点 A、 B的报文摘要数据, 比较从测量点 A、 B 收到的报文摘要数据, 匹配流标识和才艮文标识, 比较时间戳, 从而获得针 对单个 ^艮文的时延。 Step 415: The collector receives the packet summary data of the measurement points A and B, compares the packet summary data received from the measurement points A and B, matches the flow identifier and the identifier, compares the time stamp, and obtains a single ^ The delay of the text.
如果一个报文标识出现在从测量点 A收集的报文摘要数据中而没有出现 在从测量点 B收集的报文摘要数据中, 则判定该报文丟失。 If a message identifier appears in the message digest data collected from measurement point A and does not appear in the message digest data collected from measurement point B, it is determined that the message is lost.
, 步驟 420、 采集器根据单个报文的时延、 和报文丟失情况, 计算在一个计 算间隔的网络性能指标, 即平均时延、 最大时延、 最小时延、 时延抖动、 丟 包率以及网络不可用性等。 Step 420: The collector calculates the network performance indicators in a calculation interval according to the delay of the single packet and the packet loss, that is, the average delay, the maximum delay, the minimum delay, the delay jitter, and the packet loss rate. And network unavailability, etc.
步骤 425、 采集器向测量控制器发送平均时延、 最大时延、 最小时延、 时 延抖动、 丟包率以及网络不可用性等性能指标。 Step 425: The collector sends performance indicators such as average delay, maximum delay, minimum delay, delay jitter, packet loss rate, and network unavailability to the measurement controller.
步驟 430、 测量控制器向用户或者其他运营商上报本网络的性能指标。 在到达测量结束时间后, 测量点停止提取目标报文。 Step 430: The measurement controller reports the performance indicators of the network to the user or other operators. After the measurement end time is reached, the measurement point stops extracting the target message.
从上可知, 采用不仅可以支持 IPv4和 IPv6格式, 同时支持微流(由 IPv4, 五元组或者 IPv6或者三元组确定一个流)和聚合流的测量。 From the above, it can be used to support not only IPv4 and IPv6 formats, but also microflow (determining a stream by IPv4, quintuple or IPv6 or triple) and aggregated flow measurements.
本发明实施例中由测量控制器分配流标识, 或者由测量点根据被测流的 流描述生成流标识, 而不是从提取的目标报文中获得, 因此, 能够支持微流
和聚合流的测量。 根据提取的 IP报文的报文头中的内容和净荷部分的内容生 成报文标识, 不仅可以支持 IPv4和 IPv6格式, 同时支持微流(由 IPv4五元 组或者 IPv6三元组确定一个流)和聚合流的测量。 发明的精神和范围。 这样, 倘若对本发明的这些修改和变型属于本发明权利 要求及其等同技术的范围之内, 则本发明也意图包含这些改动和变型在内。
In the embodiment of the present invention, the flow identifier is allocated by the measurement controller, or the flow point identifier is generated by the measurement point according to the flow description of the measured flow, instead of being obtained from the extracted target message, and therefore, the micro flow can be supported. And measurement of aggregated flows. The packet identifier is generated according to the content in the header of the extracted IP packet and the content of the payload portion, and not only supports the IPv4 and IPv6 formats, but also supports the microflow (determining a flow by the IPv4 quintuple or the IPv6 triplet) And the measurement of the aggregate flow. The spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and the modifications of the invention
Claims
1、 一种网络性能测量方法, 其特征在于, 包括如下步驟: A network performance measurement method, comprising the steps of:
根据网络性能测量请求为被测流分配流标识, 并将流标识和被测流的流 描述配置到测量点; Assigning a flow identifier to the measured flow according to the network performance measurement request, and configuring the flow identifier and the flow description of the measured flow to the measurement point;
测量点才艮据流描述提取属于被测流的报文, 生成包含报文标识和所述分 配的流标识的报文摘要数据并上报; 以及 The measurement point extracts the packet belonging to the measured flow according to the flow description, and generates and reports the packet summary data including the packet identifier and the assigned flow identifier;
根据报文摘要数据计算网络性能指标。 Calculate network performance indicators based on packet summary data.
2、 如权利要求 1所述的方法, 其特征在于, 测量点在提取报文后进一步 附着时间戳信息, 并将该时间戳信息携带在报文摘要数据中。 2. The method according to claim 1, wherein the measurement point further attaches timestamp information after extracting the message, and carries the timestamp information in the message summary data.
3、 如权利要求 1所述的方法, 其特征在于, 测量点根据报文头中的内容 生成报文标识; 或者, 根据报文头中的内容和净荷部分的内容生成报文标识。 3. The method according to claim 1, wherein the measurement point generates a message identifier according to the content in the packet header; or generates a message identifier according to the content in the packet header and the content of the payload portion.
4、 如权利要求 3所述的方法, 其特征在于, 对于 IPv4格式报文, 所述报 文头中的内容包括报文头中的标识符和分片偏移量字段内容, 以及协议类型、 源地址和目的地址中的部分字段或全部字段的内容。 The method according to claim 3, wherein, in the IPv4 format message, the content in the packet header includes an identifier and a fragment offset field content in the packet header, and a protocol type, The content of some or all of the source and destination addresses.
5、 如权利要求 3所述的方法, 其特征在于, 对于 IPv6格式报文, 所述报 文头中的内容包括报文头中的净荷长度、 下一个头部、 源地址和目的地址字 段的内容。 The method according to claim 3, wherein, in the IPv6 format message, the content in the packet header includes a payload length, a next header, a source address, and a destination address field in the packet header. Content.
6、 如权利要求 3、 4或 5所述的方法, 其特征在于, 将报文头中的内容 或将报文头中的内容和净荷部分的内容经过函数运算后生成报文标识。 The method according to claim 3, 4 or 5, characterized in that the content in the message header or the content of the content in the header and the content of the payload portion are subjected to a function operation to generate a message identifier.
7、 如权利要求 1所述的方法, 其特征在于, 被测流的流描述为下列描述 之一: 7. The method of claim 1 wherein the stream of the stream being measured is described as one of the following descriptions:
( 1 ) 源 IP地址、 目的 IP地址、 源端口号、 目的端口号和协议类型; (1) source IP address, destination IP address, source port number, destination port number, and protocol type;
( 2 ) 源网络地址、 目的网络地址和差分编码点; (2) source network address, destination network address, and differential code point;
( 3 ) 源 IP地址、 目的 Π>地址和流标签; (3) source IP address, destination Π>address and flow label;
( 4 ) 源网络地址、 目的网络地址、 流量类型; (4) source network address, destination network address, and traffic type;
( 5 ) MPLS标签。
(5) MPLS label.
8、 一种网络性能测量方法, 其特征在于, 包括如下步骤: 8. A network performance measurement method, comprising the steps of:
根据网络性能测量将被测流的流描述配置到测量点; Configuring a flow description of the measured flow to the measurement point according to the network performance measurement;
测量点 据流描述提取属于被测流的报文, 根据所述流描述生成流标识, 以及生成包含该流标识和报文标识的报文摘要数据并上报; 以及 The measurement point is used to extract a packet that belongs to the measured flow, and generates a flow identifier according to the flow description, and generates a message summary data including the flow identifier and the packet identifier, and reports the data;
根据报文摘要数据计算网络性能指标。 Calculate network performance indicators based on packet summary data.
9、 如权利要求 8所述的方法, 其特征在于, 测量点在提取报文后进一步 附着时间戳信息, 并将该时间戳信息携带在报文摘要数据中。 9. The method according to claim 8, wherein the measurement point further attaches timestamp information after extracting the message, and carries the timestamp information in the message summary data.
10、 如权利要求 8 所述的方法, 其特征在于, 将流描述中字段的内容经 函数运算后生成流标识。 10. The method according to claim 8, wherein the content of the field in the stream description is subjected to a function operation to generate a stream identifier.
11、 如权利要求 8 所述的方法, 其特征在于, 测量点根据报文头中的内 容生成报文标识; 或者, 根据报文头中的内容和净荷部分的内容生成报文标 识。 The method according to claim 8, wherein the measurement point generates a message identifier according to the content in the message header; or, the message identifier is generated according to the content in the message header and the content of the payload portion.
12、 如权利要求 11所述的方法, 其特征在于, 对于 IPv4格式报文, 所述 报文头中的内容包括艮文头中的标识符和分片偏移量字段内容, 以及协议类 型、 源地址和目的地址中的部分字段或全部字段的内容。 The method according to claim 11, wherein, for the IPv4 format message, the content in the packet header includes an identifier and a fragment offset field content in the header, and a protocol type, The content of some or all of the source and destination addresses.
13、 如权利要求 11所述的方法, 其特征在于, 对于 IPv6格式报文, 所述 报文头中的内容包括艮文头中的净荷长度、 下一个头部、 源地址和目的地址 字段的内容。 The method according to claim 11, wherein, for the IPv6 format message, the content in the packet header includes a payload length, a next header, a source address, and a destination address field in the header. Content.
14、 如权利要求 11、 12或 13所述的方法, 其特征在于, 将报文头中的 内容或将才艮文头中的内容和净荷部分的内容经过函数运算后生成报文标识。 The method according to claim 11, 12 or 13, wherein the content of the message header or the content of the content and the payload portion of the header is subjected to a function operation to generate a message identifier.
15、 如权利要求 8 所述的方法, 其特征在于, 被测流的流描述为下列描 述之一: 15. The method of claim 8 wherein the stream of the stream being measured is described as one of the following descriptions:
( 1 ) 源 IP地址、 目的 IP地址、 源端口号、 目的端口号和协议类型; (1) source IP address, destination IP address, source port number, destination port number, and protocol type;
( 2 ) 源网络地址、 目的网络地址和差分编码点; (2) source network address, destination network address, and differential code point;
( 3 ) 源 IP地址、 目的 IP地址和流标签; (3) source IP address, destination IP address, and flow label;
( 4 ) 源网絡地址、 目的网络地址、 流量类型; (4) source network address, destination network address, and traffic type;
( 5 ) MPLS标签。
(5) MPLS label.
16、 一种网络性能测量系统, 其特征在于, 包括: 测量控制器、 釆集器 和测量装置; 16. A network performance measurement system, comprising: a measurement controller, a collector, and a measurement device;
所述测量控制器根据网络性能测量请求为被测流分配流标识, 并将流标 识和被测流的流描述配置到测量装置; The measurement controller allocates a flow identifier for the measured flow according to the network performance measurement request, and configures the flow identifier and the flow description of the measured flow to the measurement device;
所述测量装置根据流描述提取属于被测流的报文, 生成包含流标识和报 文标识的报文摘要数据并上报所述采集器; The measuring device extracts the packet belonging to the measured flow according to the flow description, generates the packet summary data including the flow identifier and the message identifier, and reports the packet summary data to the collector;
所述采集器根据上报的报文摘要数据计算网络性能指标, 并发送给测量 控制器。 The collector calculates a network performance indicator according to the reported message summary data, and sends the network performance indicator to the measurement controller.
17、 如权利要求 16所述的系统, 其特征在于, 测量装置根据报文头中的 内容生成报文标识; 或者, 根据报文头中的内容和净荷部分的内容生成报文 标识。 17. The system according to claim 16, wherein the measuring device generates a message identifier according to the content in the packet header; or generates a message identifier according to the content in the packet header and the content of the payload portion.
18、 一种网络性能测量系统, 其特征在于, 包括: 测量控制器、 采集器 , 和测量装置; 18. A network performance measurement system, comprising: a measurement controller, a collector, and a measurement device;
所述测量控制器根据网络性能测量请求将被测流的流描述配置到测量 点; The measurement controller configures a flow description of the measured flow to the measurement point according to the network performance measurement request;
所述测量装置根据流描述提取属于被测流的报文和根据所述流描述生成 流标识 , 以及生成包含流标识和报文标识的报文摘要数据并上报所述采集器; ' 所述采集器根据上艮的报文摘要数据计算网络性能指标, 并发送给测量 控制器。 The measuring device extracts a packet belonging to the measured flow according to the flow description, generates a flow identifier according to the flow description, and generates message summary data including the flow identifier and the packet identifier, and reports the packet summary data to the collector; The network calculates the network performance indicator based on the message digest data of the upper node and sends it to the measurement controller.
19、 如权利要求 18所述的系统, 其特征在于, 测量装置根据报文头中的 19. The system of claim 18, wherein the measuring device is based on a header in the message header
■ 内容生成报文标识; 或者, 根据报文头中的内容和净荷部分的内容生成报文 标识。
■ The content generation message identifier; or, the message identifier is generated according to the content in the header and the content of the payload portion.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200610072163.6 | 2006-04-14 | ||
CN2006100721636A CN101056215B (en) | 2006-04-14 | 2006-04-14 | A network performance measurement method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2007118396A1 true WO2007118396A1 (en) | 2007-10-25 |
Family
ID=38609050
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2007/000503 WO2007118396A1 (en) | 2006-04-14 | 2007-02-13 | A method and system for network performance measurement |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN101056215B (en) |
WO (1) | WO2007118396A1 (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016053693A1 (en) | 2014-09-30 | 2016-04-07 | Level 3 Communications, Llc | Sampling packets to measure network performance |
CN112363900A (en) * | 2020-11-16 | 2021-02-12 | 中国科学院计算技术研究所 | Host inner container network performance analysis method and system based on pile insertion |
CN113644998A (en) * | 2021-08-23 | 2021-11-12 | 烽火通信科技股份有限公司 | IOAM time delay measuring method and device for 5G network |
US11784895B2 (en) | 2018-12-04 | 2023-10-10 | Telecom Italia S.P.A. | Performance measurement in a packet-switched communication network |
Families Citing this family (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2685667B1 (en) | 2011-04-25 | 2015-04-08 | Huawei Technologies Co., Ltd. | Measurement method, apparatus and system based on labels |
WO2013174417A1 (en) * | 2012-05-22 | 2013-11-28 | Telecom Italia S.P.A. | Time measurement on a packet flow in a packet-switched communication network |
CN102685014B (en) * | 2012-05-29 | 2014-06-25 | 华为技术有限公司 | Method for measuring performance index of service flow and sending terminal equipment |
CN102946330B (en) | 2012-09-29 | 2017-03-15 | 华为技术有限公司 | network packet loss measuring method, device and system |
CN103109501B (en) | 2012-09-29 | 2016-11-09 | 华为技术有限公司 | Network delay measuring method, device and system |
CN102916856B (en) * | 2012-10-30 | 2015-11-11 | 中国工商银行股份有限公司 | A kind of application oriented network flow monitoring method, Apparatus and system |
CN103095521B (en) * | 2012-12-18 | 2016-03-30 | 华为技术有限公司 | The control method of flow detection, system, device, controller and checkout equipment |
CN103997429B (en) * | 2014-05-04 | 2017-04-12 | 中国科学院计算技术研究所 | Measurement system for network passive performance and method thereof |
CN105791128B (en) * | 2016-01-28 | 2018-12-14 | 东南大学 | A kind of IP packet receiving handling method and device |
CN105978751B (en) * | 2016-05-10 | 2019-04-02 | 中国航空无线电电子研究所 | A kind of UAV ground control station's delay character evaluation system |
CN106375066B (en) * | 2016-09-06 | 2019-07-23 | 北京锦鸿希电信息技术股份有限公司 | Message forming method and device of the vehicle from sensory perceptual system |
CN106506399B (en) * | 2016-11-03 | 2019-10-25 | 盛科网络(苏州)有限公司 | Realize the method, apparatus and data exchange chip of MFP |
CN109981457B (en) * | 2017-12-27 | 2021-09-07 | 华为技术有限公司 | A packet processing method, network node and system |
CN114598642A (en) | 2018-06-25 | 2022-06-07 | 华为技术有限公司 | Method and network node for sending network performance parameters, calculating network performance |
CN110943877B (en) * | 2018-09-21 | 2022-02-22 | 华为技术有限公司 | Network state measuring method, equipment and system |
CN112532468B (en) * | 2019-09-19 | 2023-03-28 | 华为技术有限公司 | Network measurement system, method, device and storage medium |
CN115208796A (en) * | 2021-03-26 | 2022-10-18 | 华为技术有限公司 | Network measurement method based on block sampling |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2005122474A1 (en) * | 2004-05-07 | 2005-12-22 | France Telecom | Performance measurement in a packet transmission network |
CN1722682A (en) * | 2004-07-15 | 2006-01-18 | 埃沃列姆公司 | Network monitoring system |
-
2006
- 2006-04-14 CN CN2006100721636A patent/CN101056215B/en not_active Expired - Fee Related
-
2007
- 2007-02-13 WO PCT/CN2007/000503 patent/WO2007118396A1/en active Application Filing
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2005122474A1 (en) * | 2004-05-07 | 2005-12-22 | France Telecom | Performance measurement in a packet transmission network |
CN1722682A (en) * | 2004-07-15 | 2006-01-18 | 埃沃列姆公司 | Network monitoring system |
Non-Patent Citations (1)
Title |
---|
LI YAO ET AL.: "Design of system for distributed real-time measurement based on IPv6, chapter 1.1 - 1.3", vol. 34, 28 February 2006, J. HUAZHONG UNIV. SCI. & TECH. (NATURE SCIENCE EDITION) * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016053693A1 (en) | 2014-09-30 | 2016-04-07 | Level 3 Communications, Llc | Sampling packets to measure network performance |
EP3202094A4 (en) * | 2014-09-30 | 2018-05-30 | Level 3 Communications, LLC | Sampling packets to measure network performance |
US10931556B2 (en) | 2014-09-30 | 2021-02-23 | Level 3 Communications, Llc | Sampling packets to measure network performance |
US11784895B2 (en) | 2018-12-04 | 2023-10-10 | Telecom Italia S.P.A. | Performance measurement in a packet-switched communication network |
CN112363900A (en) * | 2020-11-16 | 2021-02-12 | 中国科学院计算技术研究所 | Host inner container network performance analysis method and system based on pile insertion |
CN113644998A (en) * | 2021-08-23 | 2021-11-12 | 烽火通信科技股份有限公司 | IOAM time delay measuring method and device for 5G network |
CN113644998B (en) * | 2021-08-23 | 2023-06-09 | 烽火通信科技股份有限公司 | IOAM time delay measuring method and device for 5G network |
Also Published As
Publication number | Publication date |
---|---|
CN101056215A (en) | 2007-10-17 |
CN101056215B (en) | 2011-04-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2007118396A1 (en) | A method and system for network performance measurement | |
Frost et al. | Packet loss and delay measurement for mpls networks | |
US8045475B2 (en) | Method and apparatus for providing availability metrics for measurement and management of ethernet services | |
US7835290B2 (en) | Method for measuring end-to-end delay in asynchronous packet transfer network, and asynchronous packet transmitter and receiver | |
US20070064611A1 (en) | Method for monitoring packet loss ratio | |
WO2007118397A1 (en) | Measuring method for network performance and system thereof | |
WO2007118398A1 (en) | Method and system for measuring network performance | |
WO2006133635A1 (en) | A method for measurting the performance parameters of the multi-protocol label switching network | |
KR20090100377A (en) | How to Monitor and Record Performance Levels in Data Networks | |
WO2011079702A1 (en) | Method, device and router for packet loss detection | |
CN110971331B (en) | Hop-by-hop delay measurement method and system | |
KR101467137B1 (en) | In-service throughput testing in distributed router/switch architectures | |
JP2004040793A (en) | METHOD FOR ESTIMATING QoS AND COMMUNICATION DEVICE | |
US11102273B2 (en) | Uplink performance management | |
JP2023514790A (en) | NETWORK PERFORMANCE DETECTION METHOD AND DEVICE, AND NETWORK DEVICE | |
US20150036510A1 (en) | Method and device for measuring ethernet performance | |
WO2007056915A1 (en) | A method for measuring mpls network performance parameter and device and system for transmitting packet | |
Aida et al. | A scalable and lightweight QoS monitoring technique combining passive and active approaches | |
US11121938B2 (en) | Performance measurement in a packet-switched communication network | |
WO2007118399A1 (en) | A method for realizing the network performance measuring and a system thereof | |
WO2007033555A1 (en) | Method, device and system for obtaining link performance parameter | |
CN101083567A (en) | Performance monitoring of frame transmission in data network oam protocols | |
WO2007028299A1 (en) | A statistic method of the media flow in the next generation netwokr (ngn) | |
CN114640636A (en) | Cloud video management method and system | |
US7701863B2 (en) | Decentralized SLS monitoring for throughput in a differentiated service environment |
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: 07710928 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 07710928 Country of ref document: EP Kind code of ref document: A1 |