Method and apparatus for the assessment and optimization of network traffic
A system for the assessment of network performance criteria, and applying this criteria to the classification of network addresses into appropriate ranges, using these ranges to consolidate performance measurements for the associated addresses, and applying these metrics toward the optimization of the network towards performance or policy objectives.
1. A method of populating a plurality of one or more groups with a plurality of one or more network addresses, the method comprising:
selecting a plurality of one or more metrics, the plurality of the one or more metrics including statistical similarity and one or more of path performance, network cost, network usage, geographical proximity, and topological proximity;
creating the plurality of one or more groups in one or more network devices, the network devices including one or more of servers, forwarding devices, and routing devices;
populating each of the plurality of one or more groups with a subset of the plurality of one or more network addresses based on at least a classification function, the classification function at least partly determined by the selected plurality of one or more metrics; and
including one or more network addresses from one or more groups of the plurality of one or more groups in a plurality of one or more routing tables distributed across the internetwork;
wherein the at least the classification function includes a plurality of one or more criteria, the plurality of one or more criteria including topological proximity and administrative policy, wherein the topological proximity includes one or more of adjacency and dependency, and the administrative policy includes one or more of address aggregation and user-defined policy.
2. The method of claim 1 , wherein at least path performance is selected.
3. The method of claim 1 , wherein at least network cost is selected.
4. The method of claim 1 , wherein at least network usage is selected.
5. The method of claim 1 , wherein at least geographical proximity is selected.
6. The method of claim 1 , wherein at least topological proximity is selected.
7. The method of claim 1 , wherein the plurality of one or more criteria includes at least topological proximity, wherein the topological proximity includes one or more of adjacency and dependency.
8. The method of claim 1 , wherein the plurality of one or more criteria includes at least administrative policy, wherein the administrative policy includes one or more of address aggregation and user-defined policy.
9. The method of claim 1 , wherein, after the populating, the plurality of one or more groups includes a group, the group including network addresses corresponding to the plurality of one or more metrics having a value in a pre-defined range.
10. The method of claim 1 , wherein, after the populating, the plurality of one or more groups includes clusters created by a clustering algorithm applied to at least one of the plurality of one or more metrics.
11. The method of claim 1 , wherein the selecting, the creating, and the classifying are performed at a central server in an autonomous system.
12. The method of claim 1 : wherein the selecting, the creating, and the classifying are performed at a plurality of one or more distributed servers distributed through a plurality of one or more autonomous systems.
13. The method of claim 1 , wherein the selecting, creating, and populating are repeated, at least partly responsive to one or more of: a change in the at least one of the plurality of one or more metrics, an expiration of a time period, and an external input.
14. The method of 13 , wherein the selecting, creating, and populating are repeated, at least partly responsive to the change in the at least one of the one or more metrics.
15. The method of 13 , wherein the selecting, creating, and populating are repeated, at least partly responsive to the expiration of a time period.
16. The method of 13 , wherein the, selecting, creating, and populating are repeated, at least partly responsive to the external input.
17. The method of 13 , wherein, prior to the change, a first plurality of one or more groups is populated, and after the change, a second plurality of one or more groups is populated, and the first plurality of one or more groups is at least partly different from the second plurality of one or more groups.
18. The method of claim 1 , further comprising:
prior to populating the plurality of groups, creating a default population for at least one of the plurality of one or more groups, wherein the default population is derived at least partly from an observed routing table.
19. The method of claim 18 , wherein the observed routing table is from an autonomous system, such that each of the plurality of network addresses is reachable by the autonomous system.
20. The method of claim 18 , wherein the default population is a subset of the observed routing are performed on an existing routing table.
21. The method of 20 , wherein after selecting, creating, and populating are performed on the existing routing table, any changes to the existing routing table are limited to combining two or more entries into one group.
22. The method of 20 , wherein after selecting, creating, and populating are performed on the existing routing table, changes to the existing routing table include dividing at least one entry into two or more groups.
23. The method of claim 1 , further comprising:
prior to populating the plurality of groups, creating a default population for at least one of the plurality of one or more groups, wherein the default population is derived at least partly from observed network traffic.
24. The method of claim 1 , wherein the selecting, creating, and populating are performed in a first autonomous system.
25. The method of claim 24 , further comprising:
prior to populating the plurality of groups, creating a default population for at least one of the plurality of one or more groups, wherein the default population is derived at least partly from an observed routing table.
26. The method of 24 , wherein the observed routing table is from the first autonomous system.
27. The method of 24 , wherein the observed routing table is from a second autonomous system.
28. The method of claim 1 , wherein at least one subset of the plurality of one or more groups includes contiguous addresses.
29. The method of claim 1 , wherein at least one subset of the plurality of one or more groups includes at least two discontiguous addresses.
30. The method of claim 1 , where the plurality of one or more network addresses includes one or more IP addresses.
31. A network system, comprising:
a plurality of one or more network devices, the plurality of network devices including one or more of servers, forwarding devices, and routing devices, the plurality of network devices populating a plurality of one or more groups with a plurality of one or more network addresses by performing:
responsive to 1) selecting a plurality of one or more metrics, the plurality of the one or more metrics including statistical similarity of network addresses and one or more of path performance, network cost, network usage, geographical proximity, and topological proximity, and 2) creating the plurality of one or more groups in one or more network devices, the network devices including one or more of servers, forwarding devices, and routing devices;
populating each of the plurality of one or more groups with a subset of the plurality of one or more network addresses based on at least a classification function, the classification function at least partly determined by the selected plurality of one or more metrics; and
including one or more network addresses from one or more groups of the plurality of one or more groups in a plurality of one or more routing tables distributed across the internetwork;
wherein the at least the classification function includes a plurality of one or more criteria, the plurality of one or more criteria including topological proximity and administrative policy, wherein the topological proximity includes one or more of adjacency and dependency, and the administrative policy includes one or more of address aggregation and user-defined policy.
32. A network system, comprising:
a plurality of one or more network devices, the plurality of network devices including one or more of servers, forwarding devices, and routing devices, the plurality of network devices populating a plurality of one or more groups with a plurality of one or more network addresses by performing:
selecting a plurality of one or more metrics, the plurality of the one or more metrics including statistical similarity between network addresses of the one or more networked devices and the plurality of one or more metrics further including one or more of path performance, network cost, network usage, geographical proximity, and topological proximity;
creating the plurality of one or more groups in one or more network devices, the network devices including one or more of servers, forwarding devices, and routing devices, populating each of the plurality of one or more groups with a subset of the plurality of one or more network addresses based on at least a classification function, the classification function at least partly determined by the selected plurality of one or more metrics; and
including one or more network addresses from one or more groups of the plurality of one or more groups in a plurality of one or more routing tables distributed across the internetwork;
wherein the at least the classification function includes a plurality of one or more criteria, the plurality of one or more criteria including topological proximity and administrative policy, wherein the topological proximity includes one or more of adjacency and dependency, and the administrative policy includes one or more of address aggregation and user-defined policy.
33. A network system, comprising:
a plurality of one or more network devices, the plurality of network devices including one or more of servers, forwarding devices, and routing devices, the plurality of network devices populating a plurality of one or more groups with a plurality of one or more network addresses by performing:
responsive to selecting a plurality of one or more metrics, the plurality of the one or more metrics comprising pre-existing address range definitions and one or more of path performance, network cost, network usage, geographical proximity, and topological proximity;
creating the plurality of one or more groups in one or more network devices, the network devices including one or more of servers, forwarding devices, and routing devices;
populating each of the plurality of one or more groups with a subset of the plurality of one or more network addresses based on at least a classification function, the classification function at least partly determined by the selected plurality of one or more metrics; and
including one or more network addresses from one or more groups of the plurality of one or more groups in a plurality of one or more routing tables distributed across the internetwork;
wherein the at least the classification function includes a plurality of one or more criteria, the plurality of one or more criteria including topological proximity and administrative policy, wherein the topological proximity includes one or more of adjacency and dependency, and the administrative policy includes one or more of address aggregation and user-defined policy.
34. A network system, comprising:
a plurality of one or more network devices, the plurality of network devices including one or more of servers, forwarding devices, and routing devices, the plurality of network devices populating a plurality of one or more groups with a plurality of one or more network addresses by performing:
selecting a plurality of one or more metrics, the plurality of the one or more metrics including statistical similarity and one or more of path performance, network cost, network usage, geographical proximity, and topological proximity;
responsive to creating the plurality of one or more groups in one or more network devices, the network devices including one or more of servers, forwarding devices, and routing devices,
populating each of the plurality of one or more groups with a subset of the plurality of one or more network addresses based on at least a classification function, the classification function at least partly determined by the selected plurality of one or more metrics; and
including one or more network addresses from one or more groups of the plurality of one or more groups in a plurality of one or more routing tables distributed across the internetwork;
wherein the at least the classification function includes a plurality of one or more criteria, the plurality of one or more criteria including topological proximity and administrative policy, wherein the topological proximity includes one or more of adjacency and dependency, and the administrative policy includes one or more of address aggregation and user-defined policy.