1. SkipNet: A Scalable Overlay Network with Practical Locality Properties Nick Harvey, Mike Jones, Stefan Saroiu, Marvin Theimer, Alec Wolman Microsoft Research University of Washington

2. Overlay Networks Overlays have achieved several goals: Overlays have achieved several goals: Scalable and decentralized infrastructure Scalable and decentralized infrastructure Uniform and random load and data distribution Uniform and random load and data distribution But, at the price of data controllability But, at the price of data controllability Data may be stored far from its users Data may be stored far from its users Data may be stored outside its domain Data may be stored outside its domain Local accesses leave local organization Local accesses leave local organization Basic trade-off: data controllability vs. data uniformity Basic trade-off: data controllability vs. data uniformity SkipNet: SkipNet: Traditional overlay functionality Traditional overlay functionality Provides an abstraction to control this trade-off: Provides an abstraction to control this trade-off: Constrained load balancing (CLB) Constrained load balancing (CLB)

3. Talk Outline Practical data locality requirements Practical data locality requirements Basic SkipNet design Basic SkipNet design SkipNet locality properties SkipNet locality properties Performance evaluation Performance evaluation Conclusions Conclusions

4. Talk Outline Practical data locality requirements Practical data locality requirements Basic SkipNet design Basic SkipNet design SkipNet locality properties SkipNet locality properties Performance evaluation Performance evaluation Conclusions Conclusions

5. Key Locality Properties and Abstraction In practice, two properties are important: In practice, two properties are important: Content Locality – ability to explicitly place data Content Locality – ability to explicitly place data Placement on a single node or on a set of nodes Placement on a single node or on a set of nodes Path Locality – ability to guarantee that local traffic remains local Path Locality – ability to guarantee that local traffic remains local One abstraction is important – CLB: One abstraction is important – CLB: SkipNet abstraction to control the trade-off SkipNet abstraction to control the trade-off Multiple DHT scopes within one single overlay Multiple DHT scopes within one single overlay

6. Practical Requirements Data Controllability: Data Controllability: Organizations want control over their own data Organizations want control over their own data Even if local data is globally available Even if local data is globally available Manageability: Manageability: Data control allows for data administration, provisioning and manageability Data control allows for data administration, provisioning and manageability Data center/cluster = constrained set of nodes Data center/cluster = constrained set of nodes CLB ensures load balance across data center/cluster CLB ensures load balance across data center/cluster

7. Practical Requirements (contd) Security: Security: Content and path locality are key building blocks for dealing with certain external attacks Content and path locality are key building blocks for dealing with certain external attacks Data availability Data availability Local data survives network partitions Local data survives network partitions Performance Performance Data can be stored near clients that use it Data can be stored near clients that use it

8. Talk Outline Practical data locality requirements Practical data locality requirements Basic SkipNet design Basic SkipNet design SkipNet locality properties SkipNet locality properties Performance evaluation Performance evaluation Conclusions Conclusions

9. SkipNet Key property: two address spaces Key property: two address spaces 1. Name ID space: nodes are sorted by their names (e.g. DNS names) 2. Numeric ID space: nodes are randomly distributed Combining both spaces achieves Combining both spaces achieves Content + Path locality Content + Path locality Other uses could emerge: range queries [AS 03] Other uses could emerge: range queries [AS 03] Scalable peer-to-peer overlay network Scalable peer-to-peer overlay network O(log N) routing performance in both spaces O(log N) routing performance in both spaces O(log N) routing state per node O(log N) routing state per node

10. SkipNet Ring Pointers at level h skip over 2 h nodes Pointers at level h skip over 2 h nodes Nodes are ordered by names Nodes are ordered by names A D M V T X Z O

11. SkipNet Ring Pointers at level h skip over 2 h nodes Pointers at level h skip over 2 h nodes Nodes are ordered by names Nodes are ordered by names A D M V T X Z O

12. SkipNet Ring Pointers at level h skip over 2 h nodes Pointers at level h skip over 2 h nodes Nodes are ordered by names Nodes are ordered by names A E F M H S Z G

13. SkipNet Global View A Level: L = 0 L = 1 L = 3 L = 2 Root Ring Ring 0 Ring 1 Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 D M O T V X Z O Z AT M X D V A T M X D V Z O O Z A T M X D V

14. SkipNet Global View A Level: L = 0 L = 1 L = 3 L = 2 Root Ring Ring 0 Ring 1 Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 D M O T V X Z O Z AT M X D V A T M X D V Z O O Z A T M X D V

15. Two Address Spaces SkipNet can route efficiently in both address spaces: SkipNet can route efficiently in both address spaces: Name ID space (e.g. DNS names) Name ID space (e.g. DNS names) Numeric ID space Numeric ID space

16. Routing by Name ID Level: L = 0 L = 1 L = 2 Example: route from A to V Example: route from A to V Simple Rule: Forward the message to node that is closest to dest, without going too far. Simple Rule: Forward the message to node that is closest to dest, without going too far. Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3 Node As Routing Table Node As Routing Table

17. Routing by Name ID Level: L = 0 L = 1 L = 2 Example: route from A to V Example: route from A to V Simple Rule: Forward the message to node that is closest to dest, without going too far. Simple Rule: Forward the message to node that is closest to dest, without going too far. Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3

18. Routing by Name ID Level: L = 0 L = 1 L = 2 Example: route from A to V Example: route from A to V Simple Rule: Forward the message to node that is closest to dest, without going too far. Simple Rule: Forward the message to node that is closest to dest, without going too far. Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3 Node Ts Routing Table Node Ts Routing Table

19. Routing by Name ID Level: L = 0 L = 1 L = 2 Example: route from A to V Example: route from A to V Simple Rule: Forward the message to node that is closest to dest, without going too far. Simple Rule: Forward the message to node that is closest to dest, without going too far. Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3 Node Ts Routing Table Node Ts Routing Table

20. Routing by Name ID Level: L = 0 L = 1 L = 2 Example: route from A to V Example: route from A to V Simple Rule: Forward the message to node that is closest to dest, without going too far. Simple Rule: Forward the message to node that is closest to dest, without going too far. Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3 Node Ts Routing Table Node Ts Routing Table

21. Routing by Name ID Level: L = 0 L = 1 L = 2 Example: route from A to V Example: route from A to V Simple Rule: Forward the message to node that is closest to dest, without going too far. Simple Rule: Forward the message to node that is closest to dest, without going too far. Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3

22. Routing by Numeric ID Provides the basic DHT primitive Provides the basic DHT primitive To store file Foo.c To store file Foo.c Hash(Foo.c) a random numeric ID Hash(Foo.c) a random numeric ID Find highest ring matching that numeric ID Find highest ring matching that numeric ID Store file on node in that ring Store file on node in that ring Log N routing efficiency Log N routing efficiency

23. DHT Example Store file Foo.c from node A Store file Foo.c from node A Hash(Foo.c) = 101… Hash(Foo.c) = 101… Route from A to V in numeric space Route from A to V in numeric space Level: L = 0 L = 1 L = 2 Ring 00 Ring 01 Ring 10 Ring 11 Ring 000 Ring 000 Ring 001 Ring 001 Ring 010 Ring 010 Ring 011 Ring 011 Ring 100 Ring 100 Ring 101 Ring 101 Ring 110 Ring 110 Ring 111 Ring 111 A Root Ring D M O T V X Z Ring 0 A M T X Ring 1 D Z V O O Z AT M X D V A T M X D V Z O L = 3 Foo.c

24. Talk Outline Practical data locality requirements Practical data locality requirements Basic SkipNet design Basic SkipNet design SkipNet locality properties SkipNet locality properties Performance evaluation Performance evaluation Conclusions Conclusions

25. Constrained Load Balancing (CLB) Multiple DHTs with differing scopes using a single SkipNet structure Multiple DHTs with differing scopes using a single SkipNet structure A result of the ability to route in both address spaces A result of the ability to route in both address spaces Divide data object names into 2 parts using the ! special character Divide data object names into 2 parts using the ! special character CLB Domain CLB Suffix CLB Domain CLB Suffix microsoft.com ! skipnet.html microsoft.com ! skipnet.html Numeric RoutingName Routing

26. CLB Example To read file com.microsoft ! skipnet.html To read file com.microsoft ! skipnet.html Route by name ID to com.microsoft Route by name ID to com.microsoft Route by numeric ID to Hash(skipnet.html) within the com.microsoft constraint Route by numeric ID to Hash(skipnet.html) within the com.microsoft constraint com.sun edu.ucb gov.irs com.microsoft skipnet. html

27. SkipNet Path Locality Organizations correspond to contiguous SkipNet segments Organizations correspond to contiguous SkipNet segments Internal routing by NameID remains internal Internal routing by NameID remains internal Nodes have left / right pointers Nodes have left / right pointers com.sun edu.ucb gov.irs com.microsoft com.microsoft.research

28. Fault Tolerance Many failures occur along organizational boundaries: Many failures occur along organizational boundaries: Gateway/firewall failure, BGP misconfig, physical network cut, … Gateway/firewall failure, BGP misconfig, physical network cut, … SkipNet handles organizational disconnect gracefully SkipNet handles organizational disconnect gracefully Results in two well-connected, partitioned SkipNets Results in two well-connected, partitioned SkipNets Efficient remerging algorithms Efficient remerging algorithms Node independent failures Node independent failures Same resiliency as systems such as Chord and Pastry Same resiliency as systems such as Chord and Pastry Similar approach to repair (Leaf Set) Similar approach to repair (Leaf Set)

29. Primary Security Benefit & Weakness + SkipNet + name access control mechanism: Content locality ensures that content stays within organization Content locality ensures that content stays within organization Path locality prevents: Path locality prevents: malicious forwarders malicious forwarders analysis of internal traffic analysis of internal traffic external tampering external tampering - Easier to target organizations: Someone creates one million nodes with name prefixes microsofa.com and microsort.com Someone creates one million nodes with name prefixes microsofa.com and microsort.com Most traffic to/from Microsoft will go through a microsofa / microsort intermediate node Most traffic to/from Microsoft will go through a microsofa / microsort intermediate node

30. Talk Outline Practical data locality requirements Practical data locality requirements Basic SkipNet design Basic SkipNet design SkipNet locality properties SkipNet locality properties Performance evaluation Performance evaluation Conclusions Conclusions

31. Methodology Packet-level, event-driven simulator: Packet-level, event-driven simulator: SkipNet implementation SkipNet implementation Basic SkipNet Basic SkipNet Full SkipNet = Basic SkipNet + network proximity Full SkipNet = Basic SkipNet + network proximity Pastry and Chord implementation Pastry and Chord implementation Uses Mercator and GT-ITM network topologies Uses Mercator and GT-ITM network topologies Experimentally evaluated: Experimentally evaluated: Name ID routing performance Name ID routing performance Tolerance to organizational disconnect Tolerance to organizational disconnect

32. Methodology Packet-level, event-driven simulator: Packet-level, event-driven simulator: SkipNet implementation SkipNet implementation Basic SkipNet Basic SkipNet Full SkipNet = Basic SkipNet + network proximity Full SkipNet = Basic SkipNet + network proximity Pastry and Chord implementation Pastry and Chord implementation Uses Mercator and GT-ITM network topologies Uses Mercator and GT-ITM network topologies Experimentally evaluated: Experimentally evaluated: Name ID routing performance Name ID routing performance Tolerance to organizational disconnect Tolerance to organizational disconnect Numeric ID routing performance Numeric ID routing performance Effectiveness of network proximity optimizations Effectiveness of network proximity optimizations Effectiveness of CLB routing optimizations Effectiveness of CLB routing optimizations

33. Routing by Name ID Performance Benefits come at no extra cost

34. Surviving Organizational Disconnect Disconnected Org Size = 15% of all nodes

35. Conclusions SkipNet : SkipNet : Traditional overlay functionality Traditional overlay functionality Explicit control of data placement Explicit control of data placement Constrained load balancing Constrained load balancing Content + Path Locality are basic ingredients to: Content + Path Locality are basic ingredients to: Data controllability Data controllability Manageability Manageability Security Security Data availability Data availability Performance Performance

36. Questions? http://research.microsoft.com/sn/Herald/