Implementation and Evaluation of A-GAP: Adaptive Monitoring With Controllable Accuracy

The authors report on the implementation and prototype evaluation of A-GAP, a protocol for continuous monitoring of network state variables, which are computed from device counters using aggregation functions, such as SUM, AVERAGE and MAX. A-GAP is designed to achieve a given monitoring accuracy with minimal overhead. A-GAP is decentralized and asynchronous to achieve robustness and scalability. The protocol incrementally computes aggregation functions inside the network and, based on a stochastic model, it dynamically configures local filters that control the overhead and accuracy. They evaluate the prototype in a testbed of 16 commercial routers, where the protocol estimates the total number of flows in the network. Local variables are read out from Netflow buffers and aggregated in real-time.