Topology Discovery Using an Address Prefix Stopping Rule Benoit Donnet Joint work with Timur Friedman Eunice 2005 – Madrid (Spain)
Context ● Network measurement ● Internet topology discovery using distributed traceroute monitors – IP interface level ● Existing tools: – Skitter (CAIDA) – TTM (RIPE NCC) – AMP (NLANR) – DIMES (Tel Aviv U.)
Scaling Problem ● More monitors means more load on – network resources – destinations ● Classical approaches either – stay small (skitter, TTM, AMP) – trace slowly (DIMES) ● Can we trace more efficiently?
Contributions ● Efficient cooperative topology discovery algorithm – Doubletree ([Sigmetrics2005]) ● Load on destinations and communication cost reduction – probing stopping rule based on CIDR address prefixes
Doubletree - Basics ● Cooperative algorithm ● Goal: avoiding paths already explored ● Exploit tree-like structure of routes in the internet – from a monitor to a set of destinations ● Backward probing (first suggested by Govindan et al.) – from a set of monitors to a destination ● Forward probing and monitor coordination
Doubletree: Monitor-rooted tree
Doubletree: Destination-rooted tree
Doubletree: Probing scheme ● Two probing schemes: – Backwards – Forwards ● Stop sets = {(interface, root)} – Local Stop Set: B = {interface} – Global Stop Set: F = {(interface, destination)} ● shared by monitors ● Doubletree starts probing at some hop h from the monitor
Limitations ● Load on destination – Potentially, all the destinations may be hit by monitors ● Communication cost – Global stop set exchange needs too high network resources – Up to 20.6MB for only 50,000 destinations
Destination-based stopping rule
Solution ● Destination addresses aggregation through the use of CIDR prefixes ● Global stop set of {(interface, prefix_destination}
Solution (2)
Results (1)
Results (2)
Results (3)
Results (4)
Conclusion ● Improvements to Doubletree – Load on destinations reduced – Communication cost reduced ● Future works – Implementation – BGP-guided topology discovery – Clustering