HP Labs Technical Reports

Measuring and Analyzing Service Levels: A Scalable Passive Approach

Asawa, Manjari

HPL-97-138

Keyword(s): passive measurements; service level agreements; service level objectives

Abstract: Internet service providers are increasingly trying to differentiate themselves in terms of service performance that they provide to their users. To measure and quantify the service performance from end- users perspective, proper metrics and measurement methods are needed. Traditional measurement techniques are not sufficient for this purpose. In this paper, we have developed a scalable service level monitoring methodology to assess user satisfaction without injecting any measurement traffic. Specifically, we suggest web throughput as a service level metric, outline possible ways to measure it and discuss advantages of passive observations of actual user activity. We further propose a statistical data analysis method that analyzes passive throughput measurements, and quantifies user satisfaction/dissatisfaction and the confidence that the provider may have in the collected data, i.e. data reliability. The proposed technique is based on the premise that the service provider is interested in continuously monitoring the service levels being offered to a majority of the users over a long enough time. We present results of a real-world experiment that demonstrates that with careful data analysis, passive measurements are effective in detecting service problems. Our experiments also indicate that 90% of the time, the results of reliable passive measurements agree with those of active measurements, without generating any additional measurement traffic. The underlying approach may also provide a communication vehicle between service sales/marketing and operations/capacity planning aspects of service provisioning.

26 Pages

Back to Index