On the constancy of internet path properties
IMW '01 Proceedings of the 1st ACM SIGCOMM Workshop on Internet Measurement
The effects of systemic packet loss on aggregate TCP flows
Proceedings of the 2002 ACM/IEEE conference on Supercomputing
QEST '06 Proceedings of the 3rd international conference on the Quantitative Evaluation of Systems
Detection and analysis of packet loss on underutilized enterprise network links
E2EMON '05 Proceedings of the End-to-End Monitoring Techniques and Services on 2005. Workshop
Phase-Type Approximations for Message Transmission Times in Web Services Reliable Messaging
SIPEW '08 Proceedings of the SPEC international workshop on Performance Evaluation: Metrics, Models and Benchmarks
PH-distributed fault models for mobile communication
Proceedings of the Winter Simulation Conference
Fault-Impact models based on delay and packet loss for IEEE 802.11g
QEST'13 Proceedings of the 10th international conference on Quantitative Evaluation of Systems
Hi-index | 0.00 |
Injection of IP packet loss is a versatile method for emulating real-world network conditions in performance studies. In order to reproduce realistic packet-loss patterns, stochastic fault-models are used. While fault-models can be derived from measurements, inappropriate implementation may introduce artifacts in the experiment process that might invalidate results. In this paper we study the effects of different fault-models in different experiment setups. We illustrate that care should be taken to select an appropriate fault-injection method for the scenario under study.