Traffic independent survivability analysis
View Patent ↗First-order effects of hypothesized fault conditions are determined by propagating discrete test packets between select nodes and noting the change of path, if any, taken by the test packet under each condition relative to the fault-free path. Tools are provided to create classes of node pairs of interest, and test packets are created only for select classes. The network is analyzed to identify fault conditions that are likely to impact system performance, and only these fault conditions are simulated. By providing a methodology for selecting classes of node pairs to test, and prioritizing the faults to simulate, a first-order survivability analysis of large networks can be performed efficiently and effectively. The efficiency of this technique is also enhanced by providing test packets that are representative of a wide range of possible source-destination combinations, and by evaluating only the source-destination combinations that may be directly affected by each fault condition.
1. A method comprising:
identifying, on a network analysis machine, a set of source-destination pairs of a modeled network,
creating, on the network analysis machine, a plurality of test packets, each test packet corresponding to a source-destination pair of the set of source-destination pairs,
simulating, on the network analysis machine, propagation of each test packet from source to destination of the corresponding source-destination pair to determine a fault-free routing path,
identifying, on the network analysis machine, a set of fault conditions to model on the modeled network,
simulating, on the network analysis machine, each fault condition to determine if any new paths for the set of source-destination pairs is caused by the fault condition,
determining, on the network analysis machine, one or more performance measures corresponding to each new path, and
providing, on the network analysis machine, one or more reports based on the performance measures of the new paths.
2. The method of claim 1 , including receiving user input to facilitate identifying the set of source-destination pairs.
3. The method of claim 2 , wherein the user input identifies interfaces of interest and the set of source-destination pairs corresponds to distinguishable address ranges between pairs of the interfaces.
4. The method of claim 1 , wherein the simulating of each fault condition includes simulating only source-destination pairs that can be directly affected by the fault condition.
5. The method of claim 1 , wherein the performance measures include a number of hops in the new path.
6. The method of claim 1 , wherein the performance measures include a cumulative delay across the new path.
7. The method of claim 1 , including determining fault-free performance measures corresponding to the fault-free path, and wherein the reports are based on differences between the fault-free performance measures and the performance measures of the new paths.
8. The method of claim 7 , including identifying, by a user, differences that are considered to be significant differences, and wherein the reports are based on the significant differences.
9. The method of claim 8 , including receiving user input to facilitate identifying the significant differences.
10. The method of claim 9 , wherein the significant differences are differences that exceed a given threshold.
11. The method of claim 8 , wherein the significant differences are differences that exceed a given threshold.
12. The method of claim 1 , wherein identifying the set of fault conditions is based on levels of utilization of elements in the modeled network among the fault-free routing paths.
13. The method of claim 1 , wherein identifying the set of fault conditions is based on likelihoods of each fault condition causing new paths for multiple source-destination pairs.
14. The method of claim 1 , wherein the new paths are stored for each source-destination pair using equivalence classes.
15. The method of claim 1 , wherein the performance measures include a histogram that is updated when a new path is caused by the failure condition.
16. The method of claim 1 , wherein the performance measures include an average that is updated when a new path is caused by the failure condition.
17. The method of claim 1 , wherein the reports are based on groups of source-destination pairs.
18. The method of claim 17 , including receiving user input to facilitate defining the groups of source-destination pairs.
19. The method of claim 1 , wherein each source-destination pair corresponds to a single test packet.
20. A system comprising:
a pair generator that is configured to identify a set of source-destination pairs of a modeled network,
a test packet generator that is configured to create a plurality of test packets, each test packet corresponding to a source-destination pair of the set of source-destination pairs,
a simulator that is configured to simulate propagation of each test packet from source to destination of the corresponding source-destination pair to determine a fault-free routing path,
a fault condition generator that is configured to identify a set of fault conditions to model on the modeled network, and
a report generator,
wherein
the simulator is configured to simulate each fault condition to determine if any new paths for the set of source-destination pairs is caused by the fault condition, and to determine one or more performance measures corresponding to each new path, and
the report generator is configured to provide one or more reports based on the performance measures of the new paths.
21. The system of claim 20 , including a system interface that is configured to receive user input to facilitate identifying the set of source-destination pairs.
22. The system of claim 21 , wherein the user input identifies interfaces of interest and the set of source-destination pairs corresponds to distinguishable address ranges between pairs of the interfaces.
23. The system of claim 20 , wherein the simulator is configured to simulate each fault condition by simulating only source-destination pairs that can be directly affected by the fault condition.
24. The system of claim 20 , wherein the performance measures include a number of hops in the new path.
25. The system of claim 20 , wherein the performance measures include a cumulative delay across the new path.
26. The system of claim 20 , wherein the simulator is configured to determine fault-free performance measures corresponding to the fault-free path, and the reports are based on differences between the fault-free performance measures and the performance measures of the new paths.
27. The system of claim 26 , wherein the report generator is configured to allow a user to identify differences that are considered to be significant differences, and wherein the reports are based on the significant differences.
28. The system of claim 27 , including a user interface that is configured to receive user input to facilitate identifying the significant differences.
29. The system of claim 28 , wherein the significant differences are differences that exceed a given threshold.
30. The system of claim 27 , wherein the significant differences are differences that exceed a given threshold.
31. The system of claim 20 , wherein the fault condition generator is configured to identify the set of fault conditions based on levels of utilization of elements in the modeled network among the fault-free routing paths.
32. The system of claim 20 , wherein the fault condition generator is configured to identify the set of fault conditions based on likelihoods of each fault condition causing new paths for multiple source-destination pairs.
33. The system of claim 20 , wherein the new paths are stored for each source-destination pair using equivalence classes.
34. The system of claim 20 , wherein the performance measures include a histogram that is updated when a new path is caused by the failure condition.
35. The system of claim 20 , wherein the performance measures include an average that is updated when a new path is caused by the failure condition.
36. The system of claim 20 , wherein the reports are based on groups of source-destination pairs.
37. The system of claim 36 , including a user interface that is configured to receive user input to facilitate defining the groups of source-destination pairs.
38. The system of claim 20 , wherein each source-destination pair corresponds to a single test packet.
39. A non-transitory computer readable medium upon which is stored a computer program that causes a computer system to:
identify a set of source-destination pairs of a modeled network,
create a plurality of test packets, each test packet corresponding to a source-destination pair of the set of source-destination pairs,
simulate propagation of each test packet from source to destination of the corresponding source-destination pair to determine a fault-free routing path,
identify a set of fault conditions to model on the modeled network,
simulate each fault condition to determine if any new paths for the set of source-destination pairs is caused by the fault condition,
determine one or more performance measures corresponding to each new path, and
provide one or more reports based on the performance measures of the new paths.
40. The medium of claim 39 , wherein the set of source-destination pairs corresponds to distinguishable address ranges between pairs of interfaces.