This article shows how to properly interpret the ICMP-echo request latency on an EX/QFX series switch.
Example:
64 bytes from 192.168.1.1: icmp_seq=0 ttl=64 time=35.929 ms <
64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=24.838 ms < 64 bytes from 192.168.1.1: icmp_seq=2 ttl=64 time=0.870 ms 64 bytes from 192.168.1.1: icmp_seq=3 ttl=64 time=0.832 ms 64 bytes from 192.168.1.1: icmp_seq=4 ttl=64 time=0.757 ms
This is expected behavior with EX and QFX because when the switch’s RVI or L3 interface is pinged, the switch processes the request in kernel software and not in hardware. EX series switches may take 1-4 ms or more depending on CPU load to reply when a ping or ICMP-echo request is sent to the EX switch.
The nature of non-preemptive scheduling in the microkernel may, at times, introduce up to tens of milliseconds of TTP/ping packet processing delay and the delay is not uniform.
When the network interface is pinged, the traffic has to cross TTP queues and the Ethernet switch.
Do not use ICMP-echo request to measure performance.
The latency measure with an ICMP-echo request does not apply to line-rate performance or transit traffic that is hardware switched.
This behavior is noticed in both the physical and logical devices, even with the minimum interface configuration.
Latency and jitter using real time performance monitoring
Topology
Traffgen1---DUT----PeerDevice--Trafficgen2
Below config is confirmed on DUT device:
Heres the test result from sample testbed which the operator can use
root# show services | display set
set services rpm probe probe1 test test1 target address 1.1.1.2
set services rpm probe probe1 test test1 probe-count 5
set services rpm probe probe1 test test1 probe-interval 10
set services rpm probe probe1 test test1 thresholds rtt 20
run show services rpm probe-results
sample test results
{master:0}[edit]
root# run show services rpm probe-results
Owner: probe1, Test: test1
Target address: 1.1.1.2, Probe type: icmp-ping, Icmp-id: 10, Test size: 5 probes
Probe results:
Response received
Probe sent time: Thu Dec 12 11:12:06 2024
Probe rcvd/timeout time: Thu Dec 12 11:12:06 2024, No hardware timestamps
Rtt: 14288 usec, Round trip jitter: 3711 usec
Round trip interarrival jitter: 5287 usec
Results over current test:
Probes sent: 3, Probes received: 3, Loss percentage: 0.000000
Measurement: Round trip time
Samples: 3, Minimum: 5689 usec, Maximum: 14288 usec, Average: 10185 usec, Peak to peak: 8599 usec, Stddev: 3521 usec, Sum: 30554 usec
Measurement: Positive round trip jitter
Samples: 2, Minimum: 3711 usec, Maximum: 4888 usec, Average: 4300 usec, Peak to peak: 1177 usec, Stddev: 588 usec, Sum: 8599 usec
Measurement: Negative round trip jitter
Samples: 1, Minimum: 3901 usec, Maximum: 3901 usec, Average: 3901 usec, Peak to peak: 0 usec, Stddev: 0 usec, Sum: 3901 usec
Results over last test:
Probes sent: 5, Probes received: 5, Loss percentage: 0.000000
Test completed on Thu Dec 12 11:11:45 2024
Samples: 5, Minimum: 6560 usec, Maximum: 9590 usec, Average: 8493 usec, Peak to peak: 3030 usec, Stddev: 1169 usec, Sum: 42467 usec
Samples: 2, Minimum: 1867 usec, Maximum: 2810 usec, Average: 2339 usec, Peak to peak: 943 usec, Stddev: 472 usec, Sum: 4677 usec
Samples: 3, Minimum: 146 usec, Maximum: 6356 usec, Average: 2668 usec, Peak to peak: 6210 usec, Stddev: 2666 usec, Sum: 8003 usec
Results over all tests:
Probes sent: 153, Probes received: 153, Loss percentage: 0.000000
Samples: 153, Minimum: 2924 usec, Maximum: 46733 usec, Average: 11011 usec, Peak to peak: 43809 usec, Stddev: 7114 usec, Sum: 1684669 usec
Samples: 81, Minimum: 107 usec, Maximum: 28520 usec, Average: 6153 usec, Peak to peak: 28413 usec, Stddev: 7171 usec, Sum: 498425 usec
Samples: 71, Minimum: 146 usec, Maximum: 37511 usec, Average: 6863 usec, Peak to peak: 37365 usec, Stddev: 8495 usec, Sum: 487297 usec
2019-07-27: Updated to include QFX series.