Burst detection for root cause analysis of communication network soft failures

1 . A method comprising:

transmitting, by a processing system including at least one processor deployed in a communication network, instructions to a plurality of source routers in the communication network to initiate a plurality of network connectivity tests, wherein each of the plurality of network connectivity tests comprises a transmission of at least one packet via the communication network;

obtaining, by the processing system via the communication network, a set of network connectivity test results, each network connectivity test result indicating a network connectivity test success or a network connectivity test failure of a network connectivity test of the plurality of network connectivity tests between a respective router pair from a plurality of routers of the communication network, wherein the plurality of routers includes the plurality of source routers;

identifying, by the processing system from the set of network connectivity test results, a burst of network connectivity results indicating network connectivity test failures between router pairs in the communication network;

identifying, by the processing system from the set of network connectivity results, a subset of network connectivity test results indicating network connectivity test failures from within a time period associated with the burst;

determining, by the processing system for each network connectivity test result in the subset of network connectivity test results, connectivity path information comprising one or more routers between a router pair of each network connectivity test result;

determining, by the processing system from the connectivity path information of the subset of network connectivity test results, a router with a highest number of occurrences; and

identifying, by the processing system, the router as a cause of a soft failure in the communication network.

2 . The method of claim 1 , wherein the identifying of the burst of network connectivity results indicating the network connectivity test failures comprises applying a burst detection algorithm implemented by the processing system.

3 . The method of claim 2 , wherein the burst detection algorithm comprises a Kleinberg burst detection algorithm.

4 . The method of claim 2 , further comprising:

identifying at least one time period in which a number of network connectivity results indicating the network connectivity test failures exceeds a threshold.

5 . The method of claim 4 , wherein the burst comprises a shorter time period within the at least one time period.

6 . The method of claim 1 , wherein the at least one packet comprises at least one probe packet.

7 . The method of claim 1 , wherein the at least one packet comprises at least one multi-protocol label switching packet.

8 . The method of claim 7 , wherein the connectivity path information for the network connectivity test for each network connectivity test result in the subset comprises a known network path.

9 . The method of claim 8 , wherein the known network path is identified by a path identifier.

10 . The method of claim 1 , wherein the determining the router with the highest number of occurrences comprises generating a plurality of vectors, each vector associated with a network connectivity test result of the set of network connectivity test results, and each vector indicating routers along a network connectivity test path.

11 . The method of claim 10 , wherein the each vector is generated via a one-hot encoding indicating whether a given router of the communication network is on the network connectivity test path.

12 . The method of claim 11 , wherein the plurality of vectors comprises a plurality of rows of a table, wherein the determining the router with the highest number of occurrences comprises counting a number of entries in respective table columns of the table, and wherein each table column is associated with a respective router of the communication network.

13 . The method of claim 1 , further comprising:

applying at least one verification technique to confirm the router as the cause of the soft failure in the communication network.

14 . The method of claim 13 , wherein the applying the at least one verification technique comprises performing at least one hop-by-hop label-switched path verification.

15 . The method of claim 1 , further comprising:

performing at least one remedial action in the communication network in response to the identifying of the router as the cause of the soft failure in the communication network.

16 . The method of claim 15 , wherein the at least one remedial action comprises at least one of:

transmitting at least one notification; or

presenting at least one visualization.

17 . The method of claim 15 , wherein the at least one remedial action comprises:

reconfiguring at least one aspect of the communication network.

18 . A non-transitory computer-readable medium storing instructions which, when executed by a processing system including processor deployed in a communication network, cause the processing system to perform operations, the operations comprising:

transmitting first instructions to a plurality of source routers in the communication network to initiate a plurality of network connectivity tests, wherein each of the plurality of network connectivity tests comprises a transmission of at least one packet via the communication network;

obtaining, via the communication network, a set of network connectivity test results, each network connectivity test result indicating a network connectivity test success or a network connectivity test failure of a network connectivity test of the plurality of network connectivity tests between a respective router pair from a plurality of routers of the communication network, wherein the plurality of routers includes the plurality of source routers;

identifying, from the set of network connectivity test results, a burst of network connectivity results indicating network connectivity test failures between router pairs in the communication network;

identifying, from the set of network connectivity results, a subset of network connectivity test results indicating network connectivity test failures from within a time period associated with the burst;

determining, for each network connectivity test result in the subset of network connectivity test results, connectivity path information comprising one or more routers between a router pair of the network connectivity test result;

determining, from the connectivity path information of the subset of network connectivity test results, a router with a highest number of occurrences; and

identifying the router as a cause of a soft failure in the communication network.

19 . An apparatus comprising:

a processing system including at least one processor; and

a computer-readable medium storing instructions which, when executed by the processing system when deployed in a communication network, cause the processing system to perform operations, the operations comprising:

transmitting first instructions to a plurality of source routers in the communication network to initiate a plurality of network connectivity tests, wherein each of the plurality of network connectivity tests comprises a transmission of at least one packet via the communication network;

obtaining, via the communication network, a set of network connectivity test results, each network connectivity test result indicating a network connectivity test success or a network connectivity test failure of a network connectivity test of the plurality of network connectivity tests between a respective router pair from a plurality of routers of the communication network, wherein the plurality of routers includes the plurality of source routers;

identifying, from the set of network connectivity test results, a burst of network connectivity results indicating network connectivity test failures between router pairs in the communication network;

identifying, from the set of network connectivity results, a subset of network connectivity test results indicating network connectivity test failures from within a time period associated with the burst;

determining, for each network connectivity test result in the subset of network connectivity test results, connectivity path information comprising one or more routers between a router pair of the network connectivity test result;

determining, from the connectivity path information of the subset of network connectivity test results, a router with a highest number of occurrences; and

identifying the router as a cause of a soft failure in the communication network.

20 . The apparatus of claim 19 , wherein the identifying of the burst of network connectivity results indicating the network connectivity test failures comprises applying a burst detection algorithm implemented by the processing system.