Adaptive precision setting for cached approximate values
SIGMOD '01 Proceedings of the 2001 ACM SIGMOD international conference on Management of data
Adaptive filters for continuous queries over distributed data streams
Proceedings of the 2003 ACM SIGMOD international conference on Management of data
Identifying elephant flows through periodically sampled packets
Proceedings of the 4th ACM SIGCOMM conference on Internet measurement
Balancing energy efficiency and quality of aggregate data in sensor networks
The VLDB Journal — The International Journal on Very Large Data Bases
TAG: a Tiny AGgregation service for Ad-Hoc sensor networks
OSDI '02 Proceedings of the 5th symposium on Operating systems design and implementationCopyright restrictions prevent ACM from being able to make the PDFs for this conference available for downloading
A robust system for accurate real-time summaries of internet traffic
SIGMETRICS '05 Proceedings of the 2005 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Design principles and algorithms for effective high-speed IP flow monitoring
Computer Communications
A-GAP: An Adaptive Protocol for Continuous Network Monitoring with Accuracy Objectives
IEEE Transactions on Network and Service Management
Hi-index | 0.00 |
In this paper, we show the feasibility of real-time flow monitoring with controllable accuracy in today's IP networks. Our approach is based on Netflow and A-GAP. A-GAP is a protocol for continuous monitoring of network state variables, which are computed from device metrics 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. We evaluate a prototype in a testbed of 16 commercial routers and provide measurements from a scenario where the protocol continuously estimates the total number of FTP flows in the network. Local flow metrics are read out from Netflow buffers and aggregated in real-time. We evaluate the prototype for the following criteria. First, the ability to effectively control the trade off between monitoring accuracy and processing overhead; second, the ability to accurately predict the distribution of the estimation error; third, the impact of a sudden change in topology on the performance of the protocol. The testbed measurements are consistent with simulation studies we performed for different topologies and network sizes, which proves the feasibility of the protocol design, and, more generally, the feasibility of effective and efficient real-time flow monitoring in large network environments.