A network traffic expansion forecasting method and device

Abstract

Embodiments of the present invention provide a method and device for forecasting network traffic expansion, which can implement accurate forecasting of capacity expansion through a large amount of data sampled at high frequencies. The method includes: a monitoring device samples network traffic at a predetermined sampling frequency to obtain link data; converts the link traffic of a link corresponding to any link identifier ID into a relative value of the link traffic; within a predetermined time period, if If the relative value of the link traffic at any sampling time is greater than the first threshold, the link traffic of the link corresponding to any link ID at any sampling time is recorded as the accumulated value; The value is converted into a capacity expansion preparation value, and the accumulated value is cleared to zero; when the capacity expansion preparation value exceeds the second threshold, a capacity expansion notice is issued to the network management or network controller.

Description

technical field [0001] Embodiments of the present invention relate to the communication field, and in particular, to a method and device for network traffic expansion forecasting. Background technique [0002] Network monitoring is transforming from a low-frequency sampling monitoring method represented by simple network management protocol (SNMP) to a high-frequency sampling monitoring method represented by telemetry. The traditional network monitoring method mostly uses the collector to obtain monitoring data through the pull mode, and the data sampling frequency is relatively low, generally once every 5 minutes. Telemetry is a remote high-speed data collection technology from physical or virtual devices. The device actively sends device data information to the collector through push mode, providing a more real-time and higher-speed data collection function. The high-frequency sampling technology represented by telemetry can realize high-frequency sampling technology wit...

AI Technical Summary

Problems solved by technology

Because low-frequency sampling cannot fully reflect the link traffic situation, the expansion forecast based on the average traffic exceeding the threshold has smoothed out many micro-bursts in the network, so the traffic exceeding the threshold of the micro-burst cannot be reflected

The expansion forecast based on the peak traffic exceeding the threshold can obtain the maximum value of the microburst, but the average traffic of the network may be very low, and the expansion is completed in advance when the network traffic is low, resulting in waste of equipment performance and investment

After telemetry and other technologies are used, the analyzer obtains a large amount of high-frequency sampling performance data of the link, and the traditional method of exceeding the threshold or peak rate to achieve expansion forecast still brings and uses low-frequency sampling such as SNMP The monitoring method realizes the same problem as the expansion forecast, and it is impossible to realize the precise expansion forecast through a large amount of high-frequency sampling data

Images