Packet loss isolation test

1. A method for isolating packet loss on a hierarchical packet network, the method comprising:

connecting a first Network Element (NE) to a second NE via a varying path traversing multiple network segments;

discovering, with the first NE, a set of segment-demarcation expect-to-echo nodes along the varying path;

ascertaining, with the first NE, a request-to-echo configuration for each node in the set;

emitting a sample size of requests-to-echo in a respective request-to-echo configuration for each node in the set at a sample rate;

receiving results of the sample size of requests-to-echo to generate a packet-loss sample;

calculating a rate-of-loss for the packet-loss sample;

accumulating multiple packet-loss samples over a reporting interval; and

calculating an accumulated rate-of-loss for the reporting interval with the multiple packet- loss samples,

wherein the emitting emits packets marked with different Class of Service (CoS)-categorizers per packet-loss sample in a round-robin fashion for the reporting interval.

2. The method of claim 1 , wherein the emitting comprises emitting from multiple locations, and the method further comprises gathering the packet-loss samples from the multiple locations; and correlating the gathered multiple packet-loss samples to a geospatial map or one of the segments along the network path.

3. The method of claim 1 , wherein the discovering comprises identifying a traceroute protocol for each node of the set.

4. The method of claim 3 , wherein a traceroute configuration for each node comprises an Internet Control Message Protocol (ICMP) port 7 echo request-to-echo, a Transmission Control Protocol (TCP) port 7 echo request-to-echo, a User Datagram Protocol (UDP) port 7 echo request-to-echo, a UDP port 1967 request-to-echo, or a combination thereof.

5. The method of claim 1 , wherein the set comprises a Modem Ping-point (MPP), an Ingress Ping-point (IPP), an Egress Ping-Point (EPP) or a combination thereof.

6. The method of claim 1 , wherein the first NE comprises a split-tunnel connection along a network path, and the discovering discovers different segment-demarcation nodes along the split-tunnel connection.

7. The method of claim 6 , further comprising discerning an onset of excessive rate of packet-loss by comparing the rate-of-loss with an onset threshold; comparing the results along the split-tunnel connection; and deducing an at-fault segment based on the comparison.

8. The method of claim 1 , wherein the request-to-echo configuration comprises a TTL-time-exceeded traceroute request-to-echo, a fragmentation-reassembly-time-exceeded traceroute request-to-echo, a Don't Fragment-Maximum Transmission Unit (DF-MTU) destination-unreachable traceroute request-to-echo, a User Datagram Protocol (UDP) packet destined for port 123 with a Time to Live (TTL) set to decrement to zero at a respective node, a Transmission Control Protocol (TCP) packet for a port other than 80 with a TTL set to decrement to zero at the respective node, an Internet Control Message Protocol (ICMP) port 7 echo request-to-echo, a TCP port 7 echo request-to-echo, a UDP port 7 echo request-to-echo, a UDP port 1967 request-to-echo, a configured protocol and port, or a combination thereof.

9. The method of claim 1 , wherein the request-to-echo configuration comprises multiple request-to-echo configurations, and

the ascertaining comprises sending, in parallel, a request-to-echo in each of the multiple request-to-echo configurations.

10. The method of claim 1 , wherein the receiving comprises tabulating packet responses and packet losses.

11. The method of claim 1 , wherein the calculating calculates a length-of-burst-loss for the packet-loss sample.

12. The method of claim 1 , further comprising deducing an at-fault segment by comparing the rate-of-loss with an onset threshold for the results.

13. The method of claim 1 , further discerning an abatement of already-onset excessive rate of packet-loss is by comparing the rate-of-loss with an abatement threshold, or an end of a reporting interval.

14. The method of claim 1 further comprising observing a misbehavior of packet traffic, with a third NE different from the first NE and the second NE; and

triggering a periodic performance of the emitting.

15. The method of claim 1 , wherein the hierarchical packet network comprises a single-CoS layer-3 network, a single-CoS WAN network, a CoS-categorized layer-3 network, a CoS-categorized WAN network, a CoS-categorized Layer-2 container network, or a combination thereof.

16. The method of claim 1 , further comprising determining the sample size and the sample rate based on a bandwidth subscription of the first NE.

17. An underlay monitor to isolate packet loss on a hierarchical packet network, the underlay monitor comprising:

a first Network Element (NE) connected to a second NE via a varying path traversing multiple network segments;

a discoverer to discover, with the first NE, a set of segment-demarcation expect-to-echo nodes along the varying path, and to ascertain, with the first NE, a request-to-echo configuration for each node in the set;

an emitter to emit a sample size of requests-to-echo in a respective request-to-echo configuration for each node in the set at a sample rate;

a receiver to receive results of the sample size of requests-to-echo to generate a packet-loss sample; and

a calculator to calculate a rate-of-loss for the packet-loss sample,

wherein the receiver accumulates multiple packet-loss samples over a reporting interval,

the calculator further calculates an accumulated rate-of-loss for the reporting interval with the multiple packet-loss samples, and

the emitting emits packets are marked with different Class of Service (CoS)-categorizers per packet-loss sample in a round-robin fashion for the reporting interval.

18. The underlay monitor of claim 17 , wherein the discoverer identifies a traceroute protocol for each node of the set.