WO2016095410A1 - 链路流量分担的方法和装置 - Google Patents
链路流量分担的方法和装置 Download PDFInfo
- Publication number
- WO2016095410A1 WO2016095410A1 PCT/CN2015/078617 CN2015078617W WO2016095410A1 WO 2016095410 A1 WO2016095410 A1 WO 2016095410A1 CN 2015078617 W CN2015078617 W CN 2015078617W WO 2016095410 A1 WO2016095410 A1 WO 2016095410A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- time interval
- traffic
- statistical
- time
- line card
- Prior art date
Links
Images
Definitions
- the present invention relates to the field of communications, and in particular, to a method and apparatus for link traffic sharing.
- FIG. 1 is a schematic diagram of a load sharing scenario of a conventional distributed system in the related art.
- an egress link of a user network adopts dual-link uplink.
- One of the links is a leased transmission link, and the user sets the priority of the leased link to the highest.
- the forwarding traffic of the device is preferentially selected.
- the leased link with the highest priority is preferentially forwarded.
- part of the traffic is distributed to the lower-priority interface.
- the invention provides a method and a device for link traffic sharing, so as to at least solve the problem that the real-time bandwidth with high precision cannot be accurately obtained in the related art.
- a method for link traffic sharing including: receiving traffic report information sent by at least one line card, where the traffic report information includes a traffic value that is statistically calculated in a statistical time interval. Obtaining a time interval of the statistical time interval in the traffic report information of the last time sent by the at least one line card in the sampling time interval, where the sampling time interval includes at least one of the statistical time An interval; a calculation time interval is obtained according to the termination time, a time period of the statistical time interval in the sampling time interval, and a preset parameter; and the flow is counted according to the statistical time interval The quantity and the statistical time interval obtain the total flow value transmitted in the calculation time interval, and perform link flow sharing according to the total flow value.
- the method before receiving the traffic report information sent by the at least one line card, the method further includes: sending a report request message to the at least one line card; receiving the at least one line card sending The traffic report information includes: receiving the traffic report information sent by the at least one line card according to the report request message.
- the calculating the time interval according to the termination time, the time period of the statistical time interval, and the preset parameter includes: passing the formula: [t last - ⁇ *T L -S, t last - ⁇ * T L ] to obtain a calculation time interval, where t last is the termination time, T L is the time period of the statistical time interval, and the preset parameters include: S is the preset calculation time, ⁇ Is the coefficient of difference between the sampling periods between line cards.
- the ⁇ is greater than or equal to 0 and less than or equal to 2.
- a device for link traffic sharing including: a receiving module, configured to receive traffic report information sent by at least one line card, where the traffic report information includes statistics in a statistical time interval. The flow rate value and the statistical time interval; the obtaining module is configured to acquire a termination time of the statistical time interval in the last time the traffic report information of the at least one line card is sent in the sampling time interval, wherein the sampling The time interval includes at least one of the statistical time intervals; the processing module is configured to obtain a calculation time interval according to the termination time, the time period of the statistical time interval in the sampling time interval, and the preset parameter; the link sharing module And obtaining, according to the traffic value that is counted in the statistical time interval, and the statistical time interval, obtaining a total traffic value that is transmitted in the calculation time interval, and performing link traffic sharing according to the total traffic value.
- the sending module is configured to send a report request message to the at least one line card; the receiving module is further configured to receive, according to the report request message sent by the sending module, the at least one line card The traffic reports information.
- the processing module is further configured to obtain a calculation time interval by using a formula: [t last - ⁇ *T L -S, t last - ⁇ *T L ], where t last is the termination time And T L is a time period of the statistical time interval, and the preset parameter includes: S is the preset calculation time, and ⁇ is a difference coefficient between sampling periods between line cards.
- the ⁇ is greater than or equal to 0 and less than or equal to 2.
- the total flow value adds the total flow value corresponding to all the times in the calculation time interval to obtain the total
- the traffic reporting information sent by the at least one line card is included, wherein the traffic reporting information includes the traffic value that is counted in the statistical time interval and the statistical time interval, and the at least one line card is obtained in the sampling time interval.
- a termination time of the statistical time interval in the traffic report information sent at a time wherein the sampling time interval includes at least one of the statistical time intervals, and according to the termination time, the time period of the statistical time interval in the sampling time interval, and
- the preset parameter is obtained by the calculation time interval, and the total flow value transmitted in the calculation time interval is obtained according to the flow value and the statistical time interval calculated in the statistical time interval, and the link flow sharing is performed according to the total flow value.
- the traffic time-carrying device carries the statistical time interval of the traffic value in the traffic reporting information reported by the link traffic sharing device, so that the link traffic sharing device accurately obtains the traffic in the statistical time interval according to the reported statistical time interval. Values, resulting in accurate, higher real-time bandwidth.
- FIG. 1 is a schematic diagram of a load sharing scenario of a conventional distributed system in the related art
- FIG. 3 is a flowchart of a method for link traffic sharing according to an embodiment of the present invention.
- FIG. 4 is a structural block diagram of a link traffic sharing device according to an embodiment of the present invention.
- FIG. 5 is a block diagram showing a preferred structure of a link traffic sharing device according to an embodiment of the present invention.
- FIG. 6 is a block diagram showing a preferred structure of a link sharing module in a link traffic sharing device according to an embodiment of the present invention
- FIG. 7 is a schematic diagram of a real-time bandwidth calculation method in a distributed system according to an embodiment of the present invention.
- FIG. 8 is a schematic diagram of a S-second average rate calculation method according to an embodiment of the present invention.
- FIG. 9 is a preferred schematic diagram of a real-time bandwidth calculation method in a distributed system according to an embodiment of the present invention.
- FIG. 10 is a schematic diagram of a 10 second real-time rate calculation method in accordance with a preferred embodiment of the present invention.
- FIG. 3 is a flowchart of a method for link traffic sharing according to an embodiment of the present invention. As shown in FIG. 3, the process includes the following steps:
- Step S302 Receive traffic report information sent by at least one line card, where the traffic report information includes a traffic value that is counted in a statistical time interval and the statistical time interval.
- Step S304 Acquire a termination time of the statistical time interval in the traffic report information that is sent by the at least one line card in the sampling time interval, where the sampling time interval includes at least one of the statistical time intervals.
- Step S306 obtaining a calculation time interval according to the termination time, the time period of the statistical time interval in the sampling time interval, and the preset parameter;
- Step S308 Obtain a total traffic value that is transmitted in the calculation time interval according to the traffic value and the statistical time interval that are counted in the statistical time interval, and perform link traffic sharing according to the total traffic value.
- the link traffic sharing device accurately obtains the statistical time interval of the flow rate value in the traffic reporting information reported by the link traffic sharing device, so that the link traffic sharing device accurately obtains the statistical time interval according to the reported statistical time interval.
- the flow value is thus accurately obtained with a higher precision real-time bandwidth.
- the link traffic sharing device before receiving the traffic report information sent by the at least one line card, sending a report request message to the at least one line card, and receiving the traffic report message sent by the at least one line card according to the report request message .
- the link traffic sharing device actively requests the line card to report the traffic report information, so as to implement the control of reporting the traffic of each line card.
- the traffic reporting information may be ⁇ [t begin , t end ], Q ⁇ , where [t begin , t end ] is the statistical time interval, and Q is the traffic value counted in the statistical time interval. Because the line card collects the traffic continuously, the end time of the statistical time interval reported by the line card (ie, t end ) is the starting time of the statistical time interval reported by the line card (ie, t begin ).
- the calculation time interval may be obtained according to the termination time, the time period of the statistical time interval, and the preset parameter, for example, by using the formula: [t last - ⁇ * T L -S, t last - ⁇ *T L ] obtains a calculation time interval, where t last is the termination time, T L is the time period of the statistical time interval, and the preset parameters include: S is the preset calculation Time, ⁇ is the difference coefficient between the sampling periods between the line cards, so that the calculation time interval determined by the above formula can more accurately count the total flow value to achieve more accurate link sharing.
- the ⁇ is greater than or equal to 0 and less than or equal to 2.
- the ⁇ can take 0;
- the statistic time interval reported by the line card has a time delay, and the ⁇ can take 2, which is only an implementation manner, which is not limited by the present invention.
- a link traffic sharing device is provided, which is used to implement the foregoing embodiments and preferred embodiments, and details are not described herein.
- the term “module” may implement a combination of software and/or hardware of a predetermined function.
- the apparatus described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware, is also possible and contemplated.
- FIG. 4 is a structural block diagram of a link traffic sharing device according to an embodiment of the present invention.
- the link traffic sharing device 40 includes: a receiving module 42, an obtaining module 44, a processing module 46, and a link sharing module 48. The device will be described below.
- the receiving module 42 is configured to receive the traffic report information sent by the at least one line card, where the traffic report information includes the traffic value that is counted in the statistical time interval and the statistical time interval;
- the obtaining module 44 is connected to the obtaining module 44, And being configured to acquire a termination time of the statistical time interval in the last time traffic report information sent by the at least one line card received by the first receiving module 42 in the sampling time interval, where the sampling time interval includes at least one of the statistical time intervals
- the processing module 46 is connected to the obtaining module 44, and is configured to obtain a calculation time interval according to the termination time acquired by the obtaining module 44, the time period of the statistical time interval in the sampling time interval, and the preset parameter;
- the module 48 is connected to the processing module 46, and is configured to obtain the total flow value transmitted in the calculation time interval according to the flow value and the statistical time interval counted in the statistical time interval, and obtain the total flow rate according to the processing module 46.
- the value is used for link traffic sharing.
- FIG. 5 is a block diagram of a preferred structure of a link traffic sharing device according to an embodiment of the present invention. As shown in FIG. 5, the device includes, in addition to all the structures shown in FIG. 4, a sending module 52. Module 52 is described.
- the sending module 52 is connected to the receiving module 42 and configured to send a report request message to the at least one line card before the receiving module 42 receives the traffic report information sent by the at least one line card; the receiving module 42 further sets The traffic report information sent by the at least one line card according to the report request message sent by the sending module 52.
- the link traffic sharing device actively requests the line card to report the traffic report information, so as to implement the control of reporting the traffic of each line card.
- the processing module 46 is further configured to obtain a calculation time interval by using the formula: [t last - ⁇ *T L -S, t last - ⁇ *T L ], where t last is the termination time.
- T L is a time period of the statistical time interval.
- the preset parameters include: S is the preset calculation time, and ⁇ is a difference coefficient between sampling periods between line cards.
- the ⁇ is greater than or equal to 0 and less than or equal to 2.
- FIG. 6 is a block diagram showing a preferred structure of the link sharing module 48 in the link traffic sharing device according to the embodiment of the present invention.
- the link sharing module 48 includes: an obtaining unit 62, a first processing unit 64, and a The second processing unit 66 will be described below with respect to the link sharing module 48.
- the start time of the interval, t start is the start time of the at least one statistical time interval; the first processing unit 64 is connected to the obtaining unit 62, and is configured to set all the times corresponding to each time in the at least one statistical time interval.
- the link traffic sharing device may be a main control board.
- the statistical value of the sampled acquisition is transmitted to the main control CPU and the inter-process message interaction often has a gap lag, and the main control receives the The flow is actually before the TA time, and the error of the flow in the TA interval needs to be eliminated.
- FIG. 7 is a schematic diagram of a real-time bandwidth calculation method in a distributed system according to an embodiment of the present invention.
- the scheme generally includes the following processing: a resident statistical agent on a line card, and a statistical source periodicity (T L )
- T L statistical source periodicity
- the statistics agent sends statistics.
- the statistics server collects the traffic according to the traffic reported on the board. , passed to the data processing module.
- the data processing module performs real-time bandwidth calculation according to the reported policy according to the reported traffic and time.
- T R is the main control sampling period
- T L is the line card sampling period.
- the method of calculating the real-time bandwidth that is, the S-second average rate, will be described below.
- the statistical agent resides on the line card. It is used to collect and store statistical data of the statistical source, and respond to the request message sent by the master. Since the statistical agent stores the statistical data, one request message may be the data accumulated by the multiple statistical sources (the period is T L ). As shown in Figure 7, both statistical sources return a response ACK to the server.
- the reported traffic data Q t carries the traffic occurrence time interval, that is, the reporting unit is ⁇ [t begin , t end ], Q t ⁇ , as shown in FIG. 7 . Since the collection of traffic is continuous, t begin returned at the N+ 1th is the Nth t end . The reported time interval is connected end to end, and t last is the last traffic update time, which is the last reported t end .
- ⁇ is 0. It means that the traffic obtained each time can be directly used for calculation, and there is no incomplete flow in a certain time interval. This is the case with a physical interface whose traffic comes from a line card.
- ⁇ is 2. That is, the flow rate of the data arriving at the data processing module on each board can be up to two T L-MAX .
- the two-wire card cycle is the same as T L-MAX , as shown in Figure 8 (c).
- the flow rate of the board 1 is updated at t last time, but the traffic Q t 2 message of the board 2 may be updated to the master later. Therefore, the traffic of the time interval covered by Q t 1 Q t 2 in the figure is less, not Used to calculate traffic. In the limit case, the time interval with less flow is 2*T L-MAX .
- the traffic in the untrusted interval may be too small to be used as a reference for current bandwidth calculation. Subsequent as the traffic continues to update, it makes sense when the time is within the confidence interval.
- the time difference reported by each statistical source is blocked by the resident statistical agent in step 1.
- the T L and the T R are allowed to be inconsistent, and the main control can configure the T R according to the current service condition, thereby effectively reducing the main control and the line.
- Inter-card message interaction solves the first problem of the traditional device; the time when the statistics agent records the traffic update is reported together with the traffic. Since the data source carries the timestamp in step S2, the time consumption of the asynchronous message delivery is eliminated; sampling The calculation methods of steps S3, S4, and S5 well eliminate the influence of statistical source cross-board on real-time bandwidth calculation in distributed systems. For example, as illustrated in FIG. 9, as shown in FIG.
- the embodiment of the present invention uses two line cards (respectively labeled as line card 1 and line card 2), and includes two statistical time intervals in one sampling time interval.
- the sampling time interval is taken as [t0, t4]
- the two statistical time intervals of the line card 1 are respectively recorded as the statistical time interval 10 and the statistical time interval 11
- the two statistical time intervals of the line card 2 are respectively recorded as
- Q1 is the flow value reported by the line card 1
- the statistical time interval 10 is taken as [t1, t2]
- the statistical time interval 10 is taken as [t2, t3]
- Q2 is the flow rate reported by the line card 2.
- the obtained calculation time interval is [t1, t2], then the link traffic is divided into The device adds the flow rate value of the line card 1 and the flow rate value of the line card 2 at each time in the calculation time interval to obtain the total transmission value of the line card 1 and the line card 2 at each time, and each time in the calculation time interval The total value of the transmissions at a time is added to obtain the total traffic value.
- the link traffic sharing device performs link traffic sharing when the total traffic value reaches the preset traffic upper limit, and reduces the load sharing weight of the interface. Excessive traffic on the interface is distributed to other interfaces. It should be noted that the two line cards and the two statistical time intervals in the sampling time interval are only examples, which is not limited by the present invention.
- the above S can take 10 seconds.
- the preferred embodiment of the present invention will be described below by taking two line cards as an example.
- the real-time rate of load sharing calculations in a distributed system typically employs a 10 second average rate.
- the link traffic sharing scheme includes the following processing:
- the link traffic sharing device may be a main control board, and the leased link port enables the dynamic load sharing function, and configures the link priority to be non-zero, sets the bandwidth over-limit alarm threshold and the preset traffic upper limit, and the configuration command is as follows: table:
- the sampling time of the main control board and line card is 10 seconds by default, and the link aggregation (Smartgroup1) port enables performance statistics by default.
- Configure the traffic of the aggregation interface of the inter-board link Collect the traffic of the source-aware link aggregation interface from line card 1 and line card 2.
- the configuration commands are as follows:
- the statistical source sends statistics of the link aggregation port to the statistical agent on the two-line board every 10 seconds, and the statistical agent performs data accumulation, and the current traffic received by the agent is recorded, and the time interval is recorded, and the starting point of the time interval is The point in time when the traffic data is updated.
- the statistics agent needs to handle the following two situations:
- the statistics source did not complete the acquisition, and the current traffic cannot be obtained at present, and the time record is not updated.
- the collected data is zero. In this case, the traffic has been generated, so the time value needs to be updated and the time record needs to be updated.
- the performance statistics on the main control board requests data from the statistical agent of the line card every 10 seconds, and the statistical agent returns the current accumulated data to the main control board (for example, it may be the main control performance statistics module of the main control board (the same chain as above) Part of the function of the road sharing module 48)), then the main control board directly synchronizes the data to the interface performance statistics process for bandwidth calculation.
- 9 is a schematic diagram of a 10 second real-time rate calculation method according to a preferred embodiment of the present invention. As shown in FIG. 10, the Nth request, the line card 1 returns ⁇ [t 0 , t 1 ], Q 1 ⁇ , and the line card 2 returns.
- the statistical processing module divides the traffic by using a granularity of 1 second.
- Q 1 , Q 1 ', Q 2 , and Q 2 ' respectively distribute traffic data according to the reported time interval to the corresponding position. If the ⁇ is 2 for the data source cross-interface, the difference between the statistical source timing reports will not exceed one statistical period.
- the above algorithm updates the average rate of 10 seconds. In terms of the delay of the traffic rate, the selection of the calculation time interval causes the data in the last 20 seconds to be unable to participate in this calculation, so the rate calculated by the above interface has a time delay of 20 seconds.
- S6 The real-time rate is synchronized to the dynamic load sharing module, and the maximum load sharing bandwidth is obtained. If the load is more than 80%, the weight of the relevant load sharing entry is adjusted, and the load sharing weight of the interface is reduced, so that excessive traffic is dispersed on the interface. Go to other interfaces.
- the flow rate value in the statistical time interval is accurately obtained according to the reported statistical time interval, so that the high-precision real-time bandwidth is accurately obtained.
- modules or steps of the present invention described above can be implemented by a general-purpose computing device that can be centralized on a single computing device or distributed across a network of multiple computing devices. Alternatively, they may be implemented by program code executable by the computing device such that they may be stored in the storage device by the computing device and, in some cases, may be different from the order herein.
- the steps shown or described are performed, or they are separately fabricated into individual integrated circuit modules, or a plurality of modules or steps thereof are fabricated as a single integrated circuit module.
- the invention is not limited to any specific combination of hardware and software.
- the traffic time reporting information of the traffic value reported by the online card to the link traffic sharing device carries the statistical time interval of the statistics, so that the link traffic sharing device counts the statistics according to the reporting time.
- the interval accurately obtains the flow value in the statistical time interval, thereby accurately obtaining the high-precision real-time bandwidth.
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
本发明公开了一种链路流量分担的方法和装置,其中,该方法包括:接收至少一个线卡发送的流量上报信息,其中,该流量上报信息包括在统计时间区间内统计的流量值和该统计时间区间;获取在采样时间区间内该至少一个线卡最后一次发送的流量上报信息中的该统计时间区间的终止时刻,其中,该采样时间区间包括至少一个该统计时间区间;根据该终止时刻、该采样时间区间内该统计时间区间的时间段,以及预设参数得到计算时间区间;根据在该统计时间区间内统计的流量值和该统计时间区间获得该计算时间区间内传输的总流量值,并根据该总流量值进行链路流量分担,通过本发明,至少解决了相关技术中无法精确地得到精准度较高的实时带宽的问题。
Description
本发明涉及通信领域,具体而言,涉及一种链路流量分担的方法和装置。
图1是相关技术中传统分布式***的负荷分担场景示意图,如图1所示,用户网络的出口链路采用双链路上行。其中一条链路为租用的传输链路,用户将租用链路的优先级设置为最高。进入设备的转发流量,优先选择优先级最高的租用链路进行转发,当优先级最高的链路接口流量达到最大门限时,再将部分流量分担到低优先级的接口上。通过动态负荷分担,可监测所有接口的实时带宽,根据用户配置的负荷分担策略,均衡分配流量。
传统分布式***中,各CPU处理能力不相同。主控集中处理业务数据,对性能数据的处理不宜占用过多的***资源,不宜频繁采集线卡流量,因此期望线卡根据主控能力周期性上报统计值,更新实时带宽。但线卡的统计往往集成到芯片中,但各单板承载业务不同,启用时间不严格一致,各单板统计采集周期不一致,采集消息到达主控有时间差,图2是相关技术中传统统计数据采集流程图,如图2所示,t1时刻只有线卡1上报流量,线卡2中TB时间区间内数据在晚些时候t2时刻才上报。
因此,在相关技术中,无法精确地得到精准度较高的实时带宽。
发明内容
本发明提供了一种链路流量分担的方法和装置,以至少解决相关技术中无法精确地得到精准度较高的实时带宽的问题。
根据本发明的一个实施例,提供了一种链路流量分担的方法,包括:接收至少一个线卡发送的流量上报信息,其中,所述流量上报信息包括在统计时间区间内统计的流量值和所述统计时间区间;获取在采样时间区间内所述至少一个线卡最后一次发送的流量上报信息中的所述统计时间区间的终止时刻,其中,所述采样时间区间包括至少一个所述统计时间区间;根据所述终止时刻、所述采样时间区间内所述统计时间区间的时间段,以及预设参数得到计算时间区间;根据在所述统计时间区间内统计的流
量值和所述统计时间区间获得所述计算时间区间内传输的总流量值,并根据所述总流量值进行链路流量分担。
在本发明实施例,在接收所述至少一个线卡发送的所述流量上报信息之前,所述方法还包括:向所述至少一个线卡发送上报请求消息;接收所述至少一个线卡发送的流量上报信息包括:接收所述至少一个线卡根据所述上报请求消息发送的所述流量上报信息。
在本发明实施例,根据所述终止时刻、所述统计时间区间的时间段和所述预设参数得到所述计算时间区间包括:通过公式:[tlast-λ*TL-S,tlast-λ*TL]得到计算时间区间,其中,tlast为所述终止时刻,TL为所述统计时间区间的时间段,所述预设参数包括:S为所述预设计算时间,λ为线卡间的采样周期之间的差异系数。
在本发明实施例,所述λ大于等于0且小于等于2。
在本发明实施例,根据在所述统计时间区间内统计的流量值和所述统计时间区间获得所述计算时间区间内传输的总流量值包括:通过公式:L=Q/(tend-tstart)获得所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,其中,L为所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,Q为所述至少一个线卡在所述至少一个统计时间区间内统计的流量值,tend为所述至少一个统计时间区间的终止时刻,tstart为所述至少一个统计时间区间的起始时刻;将所述至少一个统计时间区间内每一时刻对应的全部线卡的流量值相加得到所述至少一个统计时间区间内每一时刻传输的流量总值;根据所述至少一个统计时间区间内每一时刻传输的流量总值将所述计算时间区间内的全部时刻对应的流量总值相加得到所述总流量值。
根据本发明的另一实施例,提供了链路流量分担的装置,包括:接收模块,设置为接收至少一个线卡发送的流量上报信息,其中,所述流量上报信息包括在统计时间区间内统计的流量值和所述统计时间区间;获取模块,设置为获取在采样时间区间内所述至少一个线卡最后一次发送的流量上报信息中的所述统计时间区间的终止时刻,其中,所述采样时间区间包括至少一个所述统计时间区间;处理模块,设置为根据所述终止时刻、所述采样时间区间内所述统计时间区间的时间段,以及预设参数得到计算时间区间;链路分担模块,设置为根据在所述统计时间区间内统计的流量值和所述统计时间区间获得所述计算时间区间内传输的总流量值,并根据所述总流量值进行链路流量分担。
在本发明实施例,发送模块,设置为向所述至少一个线卡发送上报请求消息;所述接收模块,还设置为接收所述至少一个线卡根据所述发送模块发送的上报请求消息发送的所述流量上报信息。
在本发明实施例,所述处理模块,还设置为通过公式:[tlast-λ*TL-S,tlast-λ*TL]得到计算时间区间,其中,tlast为所述终止时刻,TL为所述统计时间区间的时间段,所述预设参数包括:S为所述预设计算时间,λ为线卡间的采样周期之间的差异系数。
在本发明实施例,所述λ大于等于0且小于等于2。
在本发明实施例,所述链路分担模块包括:获得单元,设置为通过公式:L=Q/(tend-tstart)获得所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,其中,L为所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,Q为所述至少一个线卡在所述至少一个统计时间区间内统计的流量值,tend为所述至少一个统计时间区间的终止时刻,tstart为所述至少一个统计时间区间的起始时刻;第一处理单元,设置为将所述至少一个统计时间区间内每一时刻对应的全部线卡的流量值相加得到所述至少一个统计时间区间内每一时刻传输的流量总值;第二处理单元,设置为根据所述至少一个统计时间区间内每一时刻传输的流量总值将所述计算时间区间内的全部时刻对应的流量总值相加得到所述总流量值。
通过本发明,采用接收至少一个线卡发送的流量上报信息,其中,该流量上报信息包括在统计时间区间内统计的流量值和该统计时间区间,获取在采样时间区间内该至少一个线卡最后一次发送的流量上报信息中的该统计时间区间的终止时刻,其中,该采样时间区间包括至少一个该统计时间区间,并根据该终止时刻、该采样时间区间内该统计时间区间的时间段,以及预设参数得到计算时间区间,并根据在该统计时间区间内统计的流量值和该统计时间区间获得该计算时间区间内传输的总流量值,并根据该总流量值进行链路流量分担。由于线卡在向链路流量分担装置上报的流量上报信息中携带统计的流量值的统计时间区间,这样,链路流量分担装置根据该上报统计时间区间精确地得到在该统计时间区间内的流量值,从而精确地得到精准度较高的实时带宽。
此处所说明的附图用来提供对本发明的进一步理解,构成本申请的一部分,本发明的示意性实施例及其说明用于解释本发明,并不构成对本发明的不当限定。在附图中:
图1是相关技术中传统分布式***的负荷分担场景示意图;
图2是相关技术中传统统计数据采集流程图;
图3是根据本发明实施例的链路流量分担的方法的流程图;
图4是根据本发明实施例的链路流量分担装置的结构框图;
图5是根据本发明实施例的链路流量分担装置的优选结构框图;
图6是根据本发明实施例的链路流量分担装置中链路分担模块的优选结构框图;
图7是根据本发明实施例的分布式***中实时带宽计算方法示意图;
图8是根据本发明实施例的S秒平均速率计算方法示意图;
图9是根据本发明实施例的分布式***中实时带宽计算方法的优选示意图;
图10是根据本发明优选实施方式的10秒实时速率计算方法示意图。
下文中将参考附图并结合实施例来详细说明本发明。需要说明的是,在不冲突的情况下,本申请中的实施例及实施例中的特征可以相互组合。
在本实施例中提供了一种链路流量分担的方法,图3是根据本发明实施例的链路流量分担的方法的流程图,如图3所示,该流程包括如下步骤:
步骤S302,接收至少一个线卡发送的流量上报信息,其中,该流量上报信息包括在统计时间区间内统计的流量值和该统计时间区间。
步骤S304,获取在采样时间区间内该至少一个线卡最后一次发送的流量上报信息中的该统计时间区间的终止时刻,其中,该采样时间区间包括至少一个该统计时间区间。
步骤S306,根据该终止时刻、该采样时间区间内该统计时间区间的时间段,以及预设参数得到计算时间区间;
步骤S308,根据在该统计时间区间内统计的流量值和该统计时间区间获得该计算时间区间内传输的总流量值,并根据该总流量值进行链路流量分担。
这样,由于线卡在向链路流量分担装置上报的流量上报信息中携带统计的流量值的统计时间区间,这样,链路流量分担装置根据该上报统计时间区间精确地得到在该统计时间区间内的流量值,从而精确地得到精准度较高的实时带宽。
在本发明实施例,在接收该至少一个线卡发送的该流量上报信息之前,向该至少一个线卡发送上报请求消息,并接收该至少一个线卡根据该上报请求消息发送的该流量上报信息。这样,链路流量分担装置主动请求线卡上报流量上报信息,以实现对各线卡流量上报的控制。
需要说明的是,该流量上报信息可以是{[tbegin,tend],Q},其中,[tbegin,tend]为该统计时间区间,Q为在该统计时间区间内统计的流量值,由于线卡对流量的采集是持续的,因此当前该线卡上报的统计时间区间的终止时刻(即tend)即为下一次该线卡上报的统计时间区间的起始时刻(即tbegin)。
在本发明实施例,根据所述终止时刻、所述统计时间区间的时间段和所述预设参数得到所述计算时间区间可以采用多种方式,例如,可以通过公式:[tlast-λ*TL-S,tlast-λ*TL]得到计算时间区间,其中,tlast为该终止时刻,TL为该统计时间区间的时间段,该预设参数包括:S为该预设计算时间,λ为线卡间的采样周期之间的差异系数,这样,通过上述公式确定的计算时间区间能够更加精确地统计总流量值,以实现更精确的链路分担。
在本发明实施例,该λ大于等于0且小于等于2。在本发明一种可能的实现方式中,若每次获取的流量都能直接用于计算,不存在某统计时间区间内流量不完全,或者只有一个线卡时,该λ可以取0;在各个线卡上报的统计时间区间存在时延时,该λ可以取2,这里只是提供一种实现方式,本发明对此不作限定。
需要说明的是,在实际应用中,该S的设置可能会过大,从而造成tlast-λ*TL-S得出的结果为负值,此时,将该tlast-λ*TL-S取0。
在根据在所述统计时间区间内统计的流量值和所述统计时间区间获得所述计算时间区间内传输的总流量值时,也可以采用多种处理方式,在此介绍一种优选的处理方式,先通过公式:L=Q/(tend-tstart)获得该至少一个线卡在该至少一个统计时间区间内每一时刻的流量值,其中,L为该至少一个线卡在该至少一个统计时间区间内每一时刻的流量值,Q为该至少一个线卡在该至少一个统计时间区间内统计的流量值,tend为该至少一个统计时间区间的终止时刻,tstart为该至少一个统计时间区间的起始时刻;之后,将该至少一个统计时间区间内每一时刻对应的全部线卡的流量值相加得到该至少一个统计时间区间内每一时刻传输的流量总值,最后,根据该至少一个统计时间区
间内每一时刻传输的流量总值将该计算时间区间内的全部时刻对应的流量总值相加得到该总流量值,这样,通过上述公式能够更精确的计算至少一个线卡在该至少一个统计时间区间内每一时刻的流量值。
在本实施例中还提供了一种链路流量分担装置,该装置用于实现上述实施例及优选实施方式,已经进行过说明的不再赘述。如以下所使用的,术语“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现,但是硬件,或者软件和硬件的组合的实现也是可能并被构想的。
图4是根据本发明实施例的链路流量分担装置的结构框图,如图4所示,该链路流量分担装置40包括:接收模块42、获取模块44、处理模块46和链路分担模块48,下面对该装置进行说明。
接收模块42,设置为接收至少一个线卡发送的流量上报信息,其中,该流量上报信息包括在统计时间区间内统计的流量值和该统计时间区间;获取模块44,连接至上述获取模块44,设置为获取在采样时间区间内该第一接收模块42接收的至少一个线卡最后一次发送的流量上报信息中的该统计时间区间的终止时刻,其中,该采样时间区间包括至少一个该统计时间区间;处理模块46,连接至上述获取模块44,设置为根据该获取模块44获取的该终止时刻、该采样时间区间内该统计时间区间的时间段,以及预设参数得到计算时间区间;链路分担模块48,连接至上述处理模块46,设置为根据在该统计时间区间内统计的流量值和该统计时间区间获得该计算时间区间内传输的总流量值,并根据该处理模块46得到的总流量值进行链路流量分担。
图5是根据本发明实施例的链路流量分担装置的优选结构框图,如图5所示,该装置除包括图4所示的所有结构外,还包括:发送模块52,下面对该发送模块52进行说明。
发送模块52,连接至上述接收模块42,设置为在该接收模块42接收该至少一个线卡发送的该流量上报信息前,向该至少一个线卡发送上报请求消息;该接收模块42,还设置为接收该至少一个线卡根据该发送模块52发送的上报请求消息发送的该流量上报信息。这样,链路流量分担装置主动请求线卡上报流量上报信息,以实现对各线卡流量上报的控制。
在本发明实施例,上述处理模块46,还设置为通过公式:[tlast-λ*TL-S,tlast-λ*TL]得到计算时间区间,其中,tlast为该终止时刻,TL为该统计时间区间的时间段,该预设参数包括:S为该预设计算时间,λ为线卡间的采样周期之间的差异系数。
在本发明实施例,该λ大于等于0且小于等于2。
图6是根据本发明实施例的链路流量分担装置中链路分担模块48的优选结构框图,如图6所示,该链路分担模块48包括:获得单元62、第一处理单元64和第二处理单元66,下面对该链路分担模块48进行说明。
获得单元62,设置为通过公式:L=Q/(tend-tstart)获得所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,其中,L为所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,Q为所述至少一个线卡在所述至少一个统计时间区间内统计的流量值,tend为所述至少一个统计时间区间的终止时刻,tstart为所述至少一个统计时间区间的起始时刻;第一处理单元64,连接至上述获得单元62,设置为将所述至少一个统计时间区间内每一时刻对应的全部线卡的流量值相加得到所述至少一个统计时间区间内每一时刻传输的流量总值;第二处理单元66,,连接至上述第一处理单元64,设置为根据所述至少一个统计时间区间内每一时刻传输的流量总值将所述计算时间区间内的全部时刻对应的流量总值相加得到所述总流量值。需要说明的是,上述链路流量分担装置可以是主控板。
针对相关技术中的上述问题,为了得到精准度较高的实时带宽,需要解决三方面的问题:首先,如何均衡主控和线卡的性能周期不一致,主控可以自定义周期取性能数据,不能和线卡耦合;其次,若某一接口数据取自多块单板,各单板统计值周期性上报的时刻不一致,在主控一次收到的流量更新只是某一块单板的,无法预估其他单板流量值和上报时间,因此计算的实时带宽偏小。最后,分布式***中,统计值在消息传递过程中会有时延,在业务量大时,采样获取的统计值传递到主控CPU和进程间消息交互往往会间隙性滞后,主控收到的流量其实是TA时间前的,TA区间的流量的误差需要消除。
基于上述对相关技术中存在的分布式***中统计周期不一致、带宽计算不精准、多统计源无法实时计算带宽的问题的分析,在本实施例中,提出了一种分布式***中实时带宽数据采集和计算的方案。图7是根据本发明实施例的分布式***中实时带宽计算方法示意图,如图7所示,该方案大致包括如下处理:线卡上驻留统计代理,统计源周期性(TL)的往统计代理发送统计数据。主控上的统计服务端,周期性(TR)的往驻留统计代理的进程请求统计结果,统计代理回统计请求给统计服务端,统计服务端根据不同单板上报的流量进行汇聚存储后,传递给数据处理模块。数据处理模块,将流量值根据上报的流量及时间按照相应的策略进行实时带宽计算。其中,图中TR为主控采样周期,TL为线卡采样周期。
下面对其中实时带宽即S秒平均速率计算的方法进行说明。
S1,线卡上驻留统计代理。用来收集并存储统计源的统计数据,同时回应主控发来的请求消息。由于统计代理存储了统计数据,一次请求消息可能是多次统计源(周期为TL)上报的数据累加。如图7中,两个统计源均给服务端回了响应ACK。
S2,上报流量数据Qt携带流量发生时间区间,即上报单元为{[tbegin,tend],Qt},如图7。由于流量的采集是持续的,因此第N+1次返回的tbegin为第N次的tend。上报的时间区间,首尾相接,tlast为最后一次流量更新时间也就是最后一次上报的tend。
S3,统计处理模块,按秒分割时间轴,将流量Qt按时间比例线性划分到相应的时间区间内。图8是根据本发明实施例的S秒平均速率计算方法示意图,如图8所示,该区间内,每秒流量qt=Qt/(tend-tbegin)如图8(a)。若t秒有多次流量上报则t秒流量:qt=qt
1+qt
2+qt
3……+qt
n。
S4,设定可用于计算速率的时间区间。本文称该区间为置信区间。考虑到各板采集周期不一致,取系数λ反映上述差异。TL-MAX取各单板上报周期最大值。则时间轴上[0,tlast-λ*TL-MAX]为置信区间,如图8(b)。
下面对λ的取值进行说明。
λ为0。表示每次获取的流量都能直接用于计算,不存在某时间区间内流量不完全。物理接口即是这样,其流量来自某一线卡。
λ为2。即各单板上报到达数据处理模块的流量其时间起点和终点最大可相差两个TL-MAX。典型的当两线卡周期均相同为TL-MAX,如图8中(c)。tlast时刻单板1的流量更新,但是单板2的流量Qt
2消息可能稍后才能更新到主控,因此图中Qt
1Qt
2覆盖的时间区间流量都是偏少的,不可用来计算流量。极限情况下流量偏少的时间区间为2*TL-MAX。
实际***中受各单板周期差异的影响,λ取值范围建议为[0,2]。
S5,取[tlast-λ*TL-MAX-S,tlast-λ*TL-MAX]时间区间内的t时刻流量计算S秒平均速率。即将S秒区间内步骤3划分的流量累加。即:
qt-1+qt-2+qt-3……+qt-s
S6,非置信区间内的流量可能是偏少的,不能作为当前带宽计算的参考。后续随着流量的持续更新,当时间处于置信区间内才有意义。
在上述实施例中,通过步骤1中驻留统计代理,屏蔽了各统计源上报的时间差,TL和TR允许不一致,主控可以根据当前业务情况配置TR,有效减少了主控与线卡间消息交互,解决了传统设备存在的第一个问题;统计代理记录流量更新的时间,与流量一同上报,由于步骤S2中数据源处携带时间戳,消除了异步消息传递的时间消耗;采样步骤S3,S4,S5的计算方式很好消除了分布式***中统计源跨板对实时带宽计算的影响。示例地,结合图9进行说明,如图9所示,本发明实施例以两个线卡(分别记为线卡1和线卡2),且一个采样时间区间内包括两个统计时间区间为例进行说明,其中,采样时间区间取[t0,t4],线卡1的两个统计时间区间分别记为统计时间区间10和统计时间区间11,线卡2的两个统计时间区间分别记为统计时间区间20和统计时间区间21,Q1为线卡1上报的流量值,统计时间区间10取[t1,t2],统计时间区间10取[t2,t3],Q2为线卡2上报的流量值,统计时间区间20取[t0,t1],统计时间区间21取[t1,t2],则链路流量分担装置按照上述公式分别计算线卡1在统计时间区间10内每个时刻的流量值、线卡1在统计时间区间11内每个时刻的流量值和线卡2在统计时间区间20内每个时刻的流量值以及线卡2在统计时间区间21内每个时刻的流量值,若得到的计算时间区间为[t1,t2],则该链路流量分担装置将该计算时间区间内每一时刻的线卡1的流量值和线卡2的流量值相加得到每一时刻线卡1和线卡2的传输总值,并将该计算时间区间内每一时刻的传输总值相加得到总流量值,这样,链路流量分担装置在确定该总流量值在达到预设流量上限时,进行链路流量分担,减少该接口的负荷分担权重,使该接口上过多的流量分散到其他接口上,需要说明的是,上述两个线卡以及采样时间区间内包括两个统计时间区间只是举例说明,本发明对此不作限定。
在本发明实施例,上述的S可以取10秒,下面以两个线卡为例对本发明优选实施方式进行说明。
在该实施例中,分布式***中负荷分担计算的实时速率一般采用10秒平均速率。设定主控和线卡均10秒采样周期。该链路流量分担方案包括如下处理:
S1,链路流量分担装置可以是主控板,租用链路端口使能动态负荷分担功能,并将链路优先级配置非0,设定带宽超限告警阈值和预设流量上限,配置指令如下表:
S2,主控板和线卡采样时间默认均为10秒,同时链路聚合(Smartgroup1)口默认使能性能统计。配置跨板链路聚合口业务,统计源感知链路聚合口流量采集自线卡1和线卡2,配置指令如下:
S3,统计源每10秒向两线板上统计代理发送链路聚合口的统计数据,并由统计代理做数据累加,统计代理收到的当前流量,记录下时间区间,时间区间的起点就是上一次流量数据更新的时间点。如{[t0,t1],Q1}等。统计代理需要处理下面两种情形:
(a)采集数据未上报,(b)统计周期内采集数据为零;
第一种情形,统计源未完成该次采集,当前无法获取确切流量,时间记录也不更新。第二种情形,采集数据为零,此情形流量已生成,因此时间值需要更新,需要更新时间记录。
S4,主控板上性能统计每10秒向线卡的统计代理请求数据,统计代理将当前累加的数据返回给主控板(例如,可以是主控板的主控性能统计模块(同上述链路分担模块48的部分功能)),之后主控板将该数据直接同步给接口性能统计进程进行带宽计算。图9是根据本发明优选实施方式的10秒实时速率计算方法示意图,如图10所示,第N次请求,线卡1返回{[t0,t1],Q1},线卡2返回{[t0',t1'],Q1'},第N+1次请求,线卡1返回{[t1,t2],Q2},线卡2返回{[t1',t2'],Q2'}。
S5,统计处理模块以1秒为粒度划分流量。Q1,Q1',Q2,Q2'分别根据上报的时间区间往相应位置均摊流量数据。对于数据源跨板接口取λ为2,即认为单板间统计源定时上报的差异不会超过一个统计周期,最后由上文算法更新10秒平均速率。流量速率的延时方面,计算时间区间的选择导致最近20秒内的数据无法参与本次计算,因此上述接口计算得到的速率存在20秒的时间延迟。
S6,实时速率同步至动态负荷分担模块,并获取最大负荷分担带宽,如果超过80%,则调整相关负荷分担表项的权重,减少该接口的负荷分担权重,使该接口上过多的流量分散到其他接口上。
通过上述优选实施方式,根据该上报统计时间区间精确地得到在该统计时间区间内的流量值,从而精确地得到精准度较高的实时带宽。
显然,本领域的技术人员应该明白,上述的本发明的各模块或各步骤可以用通用的计算装置来实现,它们可以集中在单个的计算装置上,或者分布在多个计算装置所组成的网络上,可选地,它们可以用计算装置可执行的程序代码来实现,从而,可以将它们存储在存储装置中由计算装置来执行,并且在某些情况下,可以以不同于此处的顺序执行所示出或描述的步骤,或者将它们分别制作成各个集成电路模块,或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样,本发明不限制于任何特定的硬件和软件结合。
以上所述仅为本发明的优选实施例而已,并不用于限制本发明,对于本领域的技术人员来说,本发明可以有各种更改和变化。凡在本发明的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。
基于本发明的实施例提供的上述技术方案,通过在线卡在向链路流量分担装置上报的流量上报信息中携带统计的流量值的统计时间区间,这样,链路流量分担装置根据该上报统计时间区间精确地得到在该统计时间区间内的流量值,从而精确地得到精准度较高的实时带宽。
Claims (10)
- 一种链路流量分担的方法,包括:接收至少一个线卡发送的流量上报信息,其中,所述流量上报信息包括在统计时间区间内统计的流量值和所述统计时间区间;获取在采样时间区间内所述至少一个线卡最后一次发送的流量上报信息中的所述统计时间区间的终止时刻,其中,所述采样时间区间包括至少一个所述统计时间区间;根据所述终止时刻、所述采样时间区间内所述统计时间区间的时间段,以及预设参数得到计算时间区间;根据在所述统计时间区间内统计的流量值和所述统计时间区间获得所述计算时间区间内传输的总流量值,并根据所述总流量值进行链路流量分担。
- 根据权利要求1所述的方法,其中,在接收所述至少一个线卡发送的所述流量上报信息之前,所述方法还包括:向所述至少一个线卡发送上报请求消息;接收所述至少一个线卡发送的流量上报信息包括:接收所述至少一个线卡根据所述上报请求消息发送的所述流量上报信息。
- 根据权利要求1所述的方法,其中,根据所述终止时刻、所述统计时间区间的时间段和所述预设参数得到所述计算时间区间包括:通过公式:[tlast-λ*TL-S,tlast-λ*TL]得到计算时间区间,其中,tlast为所述终止时刻,TL为所述统计时间区间的时间段,所述预设参数包括:S为所述预设计算时间,λ为线卡间的采样周期之间的差异系数。
- 根据权利要求3所述的方法,其中,所述λ大于等于0且小于等于2。
- 根据权利要求1至4任一项所述的方法,其中,根据在所述统计时间区间内统计的流量值和所述统计时间区间获得所述计算时间区间内传输的总流量值包括:通过公式:L=Q/(tend-tstart)获得所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,其中,L为所述至少一个线卡在所述至少一个统 计时间区间内每一时刻的流量值,Q为所述至少一个线卡在所述至少一个统计时间区间内统计的流量值,tend为所述至少一个统计时间区间的终止时刻,tstart为所述至少一个统计时间区间的起始时刻;将所述至少一个统计时间区间内每一时刻对应的全部线卡的流量值相加得到所述至少一个统计时间区间内每一时刻传输的流量总值;根据所述至少一个统计时间区间内每一时刻传输的流量总值将所述计算时间区间内的全部时刻对应的流量总值相加得到所述总流量值。
- 一种链路流量分担的装置,包括:接收模块,设置为接收至少一个线卡发送的流量上报信息,其中,所述流量上报信息包括在统计时间区间内统计的流量值和所述统计时间区间;获取模块,设置为获取在采样时间区间内所述至少一个线卡最后一次发送的流量上报信息中的所述统计时间区间的终止时刻,其中,所述采样时间区间包括至少一个所述统计时间区间;处理模块,设置为根据所述终止时刻、所述采样时间区间内所述统计时间区间的时间段,以及预设参数得到计算时间区间;链路分担模块,设置为根据在所述统计时间区间内统计的流量值和所述统计时间区间获得所述计算时间区间内传输的总流量值,并根据所述总流量值进行链路流量分担。
- 根据权利要求6所述的装置,其中,所述装置还包括:发送模块,设置为向所述至少一个线卡发送上报请求消息;所述接收模块,还设置为接收所述至少一个线卡根据所述发送模块发送的上报请求消息发送的所述流量上报信息。
- 根据权利要求6所述的装置,其中,所述处理模块,还设置为通过公式:[tlast-λ*TL-S,tlast-λ*TL]得到计算时间区间,其中,tlast为所述终止时刻,TL为所述统计时间区间的时间段,所述预设参数包括:S为所述预设计算时间,λ为线卡间的采样周期之间的差异系数。
- 根据权利要求8所述的装置,其中,所述λ大于等于0且小于等于2。
- 根据权利要求6至9任一项所述的装置,其中,所述链路分担模块包括:获得单元,设置为通过公式:L=Q/(tend-tstart)获得所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,其中,L为所述至少一个线卡在所述至少一个统计时间区间内每一时刻的流量值,Q为所述至少一个线卡在所述至少一个统计时间区间内统计的流量值,tend为所述至少一个统计时间区间的终止时刻,tstart为所述至少一个统计时间区间的起始时刻;第一处理单元,设置为将所述至少一个统计时间区间内每一时刻对应的全部线卡的流量值相加得到所述至少一个统计时间区间内每一时刻传输的流量总值;第二处理单元,设置为根据所述至少一个统计时间区间内每一时刻传输的流量总值将所述计算时间区间内的全部时刻对应的流量总值相加得到所述总流量值。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410802351.4 | 2014-12-19 | ||
CN201410802351.4A CN105791144A (zh) | 2014-12-19 | 2014-12-19 | 链路流量分担的方法和装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016095410A1 true WO2016095410A1 (zh) | 2016-06-23 |
Family
ID=56125763
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2015/078617 WO2016095410A1 (zh) | 2014-12-19 | 2015-05-08 | 链路流量分担的方法和装置 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN105791144A (zh) |
WO (1) | WO2016095410A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110011926A (zh) * | 2019-03-07 | 2019-07-12 | 新华三技术有限公司 | 一种调整报文发送时间的方法、装置、设备及存储介质 |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107733805B (zh) * | 2016-08-12 | 2021-04-16 | 腾讯科技(深圳)有限公司 | 业务负载调度方法和装置 |
CN111211987B (zh) * | 2019-11-29 | 2021-07-27 | 清华大学 | 网络中流量动态调整方法、***、电子设备及存储介质 |
CN113347107B (zh) * | 2020-03-02 | 2022-10-14 | ***通信集团浙江有限公司 | 基于上行报文的流量调度方法、装置及计算设备 |
CN114070707A (zh) * | 2020-11-10 | 2022-02-18 | 北京市天元网络技术股份有限公司 | 一种互联网性能监控方法及*** |
CN112714028A (zh) * | 2021-01-29 | 2021-04-27 | 紫光云技术有限公司 | 一种负载均衡slb流量、数据包参数的监控方法 |
CN113328906B (zh) * | 2021-04-22 | 2023-01-06 | 成都欧珀通信科技有限公司 | 一种流量实时监控方法、装置、存储介质及电子设备 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020103631A1 (en) * | 2000-04-21 | 2002-08-01 | Anja Feldmann | Traffic engineering system and method |
CN1798096A (zh) * | 2004-12-28 | 2006-07-05 | 杭州华为三康技术有限公司 | 在路由器或交换机上进行多出口流量管理的方法和设备 |
US20060176888A1 (en) * | 1998-09-17 | 2006-08-10 | Mcnamara Tod W | High density network topology |
CN102143042A (zh) * | 2010-07-09 | 2011-08-03 | 华为技术有限公司 | 虚拟集群路由器***及其流量分担方法、控制器和子路由器 |
CN102932268A (zh) * | 2012-11-05 | 2013-02-13 | 杭州迪普科技有限公司 | 一种数据报文转发方法和装置 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1984119B (zh) * | 2005-12-13 | 2010-06-23 | 中兴通讯股份有限公司 | 一种利用时分技术实现流量控制的方法 |
US7768926B2 (en) * | 2006-03-09 | 2010-08-03 | Firetide, Inc. | Effective bandwidth path metric and path computation method for wireless mesh networks with wired links |
CN101257415B (zh) * | 2008-01-18 | 2010-08-11 | 东南大学 | 基于固定存储空间的网络流实时自适应测量方法 |
CN102237957B (zh) * | 2010-05-07 | 2013-07-24 | 电信科学技术研究院 | 一种信道状态信息的上报方法、***及装置 |
-
2014
- 2014-12-19 CN CN201410802351.4A patent/CN105791144A/zh active Pending
-
2015
- 2015-05-08 WO PCT/CN2015/078617 patent/WO2016095410A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060176888A1 (en) * | 1998-09-17 | 2006-08-10 | Mcnamara Tod W | High density network topology |
US20020103631A1 (en) * | 2000-04-21 | 2002-08-01 | Anja Feldmann | Traffic engineering system and method |
CN1798096A (zh) * | 2004-12-28 | 2006-07-05 | 杭州华为三康技术有限公司 | 在路由器或交换机上进行多出口流量管理的方法和设备 |
CN102143042A (zh) * | 2010-07-09 | 2011-08-03 | 华为技术有限公司 | 虚拟集群路由器***及其流量分担方法、控制器和子路由器 |
CN102932268A (zh) * | 2012-11-05 | 2013-02-13 | 杭州迪普科技有限公司 | 一种数据报文转发方法和装置 |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110011926A (zh) * | 2019-03-07 | 2019-07-12 | 新华三技术有限公司 | 一种调整报文发送时间的方法、装置、设备及存储介质 |
Also Published As
Publication number | Publication date |
---|---|
CN105791144A (zh) | 2016-07-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2016095410A1 (zh) | 链路流量分担的方法和装置 | |
KR102385762B1 (ko) | 속도 최적화된 정체 관리 | |
US8730819B2 (en) | Flexible network measurement | |
CN110535965A (zh) | 一种数据处理方法及装置、存储介质 | |
JPWO2018180369A1 (ja) | センサネットワークシステム | |
CN101272290B (zh) | Ip网络中路径拥塞状态的测量方法和测量装置 | |
US20200244574A1 (en) | Path computation method, message responding method, and related device | |
US9350669B2 (en) | Network apparatus, performance control method, and network system | |
CN110620699B (zh) | 消息到达率确定方法、装置、设备和计算机可读存储介质 | |
US11641245B2 (en) | Timestamp confidence level | |
CN109828843A (zh) | 一种计算节点间数据传输的方法、***及电子设备 | |
US20170163509A1 (en) | Inter-node distance metric method and system | |
Wei et al. | Shared bottleneck detection based on congestion interval variance measurement | |
CN104580322A (zh) | 一种分布式数据流处理方法及装置 | |
CN113328906B (zh) | 一种流量实时监控方法、装置、存储介质及电子设备 | |
WO2017096837A1 (zh) | 节点间距离的度量方法及*** | |
CN105591790B (zh) | 数据通信连接池管理装置 | |
CN109787861B (zh) | 网络数据延迟控制方法 | |
US20100208732A1 (en) | Controlling Packet Transmission | |
CN115801639B (zh) | 一种带宽探测方法、装置、电子设备及存储介质 | |
US20220360534A1 (en) | Communication equipment, communication methods and programs | |
CN114024913B (zh) | 一种网络性能优化方法、装置、设备以及存储介质 | |
CN104270466A (zh) | 数据上报方法及相关设备 | |
WO2016165471A1 (zh) | Hadoop集群组件指标的采集方法及装置 | |
JP5923914B2 (ja) | 網状態推定装置及び網状態推定プログラム |
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: 15868923 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: 15868923 Country of ref document: EP Kind code of ref document: A1 |