Performance impact analysis of network change

摘要

A network server analyzes a change in the network, including performing a machine-learning analysis of an extrapolation space. The server accesses observed data from multiple counters that each record samples for a metric in the network. The server performs a CART (classification and regression tree) analysis of the observed data to select the counters whose metrics affect a target network performance, such as latency. The server estimates an extrapolation space based on the observed data for the selected counters. The server then performs a machine-learning analysis of the extrapolation space based on a kriging model of the selected counters.

主权项

1. A method for analyzing a change in a network, comprising:
accessing multiple counters of a network system to obtain observed data, each counter of the multiple counters recording samples for a metric in the network; performing a classification and regression tree (CART) analysis of the obtained observed data to select a subset of the multiple counters whose metrics are related to a target network performance; extrapolating with a kriging-based analysis of the subset of the multiple counters selected from the CART analysis, wherein said extrapolating yields data for the selected subset of the multiple counters beyond the observed data; and indicating whether a criterion of a to-be-provisioned workload will be satisfied based, at least in part, on the data yielded from said extrapolating and whether a criterion of an already provisioned workload will be violated based, at least in part on the data yielded from the extrapolating.