A value of "0" for the object pingResultsAverageRtt implies that no round-trip-time(RTT) has been received.
While SNMP-Agent polls RTT related standard OIDs from a Junos router such as 'pingResultsAverageRtt', the Junos router responds with zero (0) value even though there is no ping failure.
The following objects may be impacted:
iso.3.6.1.2.1.80.1.3.1.4 -> pingResultsMinRtt iso.3.6.1.2.1.80.1.3.1.5 -> pingResultsMaxRtt iso.3.6.1.2.1.80.1.3.1.6 -> pingResultsAverageRtt iso.3.6.1.2.1.80.1.3.1.7 -> pingResultsProbeResponses iso.3.6.1.2.1.80.1.3.1.9 -> pingResultsRttSumOfSquares
Configuration:
user@router# show services rpm probe 405 { test 405 { probe-type icmp-ping; target address 10.1.5.2; probe-interval 3; test-interval 10; thresholds { successive-loss 3; } traps [ probe-failure test-failure ]; } } user@router# show snmp view public { oid .1; } community public { view public; authorization read-only; clients { 0.0.0.0/0; } } traceoptions { file snmp.log size 50m files 10; flag all; }
SNMP-Agent walks 'pingResultsTable' to the router continually, and may get zero value for pingResultsAverageRtt occasionally.
snmpd traceoption:
Sep 4 16:49:46.850558 snmpd[217ac5f6] >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Sep 4 16:49:46.852174 AgentX subagent: /var/run/rmopd-17, PDU request id: 158026, trans id: 158025, rsp(us): 1440 Sep 4 16:49:46.852312 snmpd[217ac5f6] <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< Sep 4 16:49:46.852349 snmpd[217ac5f6] <<< Get-Response Sep 4 16:49:46.852385 snmpd[217ac5f6] <<< Source: 10.85.248.20 Sep 4 16:49:46.852419 snmpd[217ac5f6] <<< Destination: 172.27.100.185 Sep 4 16:49:46.852449 snmpd[217ac5f6] <<< Version: SNMPv2 Sep 4 16:49:46.852480 snmpd[217ac5f6] <<< Request_id: 0x217ac5f6 Sep 4 16:49:46.852513 snmpd[217ac5f6] <<< Community: public Sep 4 16:49:46.852544 snmpd[217ac5f6] <<< Error: status=0 / vb_index=0 Sep 4 16:49:46.852597 snmpd[217ac5f6] <<< OID : pingResultsAverageRtt.3.52.48.53 Sep 4 16:49:46.852636 snmpd[217ac5f6] <<< .3.52.48.53 Sep 4 16:49:46.852668 snmpd[217ac5f6] <<< type : Gauge Sep 4 16:49:46.852701 snmpd[217ac5f6] <<< value: 0 Sep 4 16:49:46.852730 snmpd[217ac5f6] <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< Sep 4 16:49:46.852834 SNMP Target: 172.27.100.185, PDU request id: 561694198, rsp(us): 2096
When checking the rpm probe-results, there is no probe failure.
user@router> show services rpm probe-results Sep 4 17:08:06 Owner: 405, Test: 405 Target address: 10.1.5.2, Probe type: icmp-ping, Icmp-id: 10, Test size: 15 probes Probe results: Response received Probe sent time: Wed Sep 4 17:08:06 2019 Probe rcvd/timeout time: Wed Sep 4 17:08:06 2019, No hardware timestamps Rtt: 1041 usec, Round trip jitter: -45 usec Round trip interarrival jitter: 365 usec Results over current test: Probes sent: 11, Probes received: 11, Loss percentage: 0.000000 Measurement: Round trip time Samples: 11, Minimum: 1029 usec, Maximum: 1711 usec, Average: 1197 usec, Peak to peak: 682 usec, Stddev: 217 usec, Sum: 13158 usec Measurement: Positive round trip jitter Samples: 5, Minimum: 18 usec, Maximum: 628 usec, Average: 177 usec, Peak to peak: 610 usec, Stddev: 230 usec, Sum: 884 usec Measurement: Negative round trip jitter Samples: 6, Minimum: 42 usec, Maximum: 384 usec, Average: 155 usec, Peak to peak: 342 usec, Stddev: 137 usec, Sum: 926 usec Results over last test: Probes sent: 15, Probes received: 15, Loss percentage: 0.000000 Test completed on Wed Sep 4 17:07:53 2019 Measurement: Round trip time Samples: 15, Minimum: 1019 usec, Maximum: 1621 usec, Average: 1118 usec, Peak to peak: 602 usec, Stddev: 194 usec, Sum: 16758 usec Measurement: Positive round trip jitter Samples: 8, Minimum: 1 usec, Maximum: 587 usec, Average: 93 usec, Peak to peak: 586 usec, Stddev: 188 usec, Sum: 740 usec Measurement: Negative round trip jitter Samples: 7, Minimum: 3 usec, Maximum: 554 usec, Average: 98 usec, Peak to peak: 551 usec, Stddev: 187 usec, Sum: 681 usec Results over all tests: Probes sent: 366206, Probes received: 366206, Loss percentage: 0.000000 Measurement: Round trip time Samples: 366206, Minimum: 807 usec, Maximum: 853578 usec, Average: 1972 usec, Peak to peak: 852771 usec, Stddev: 4083 usec, Sum: 722067893 usec Measurement: Positive round trip jitter Samples: 177989, Minimum: 0 usec, Maximum: 852534 usec, Average: 1455 usec, Peak to peak: 852534 usec, Stddev: 5608 usec, Sum: 258803972 usec Measurement: Negative round trip jitter Samples: 188216, Minimum: 1 usec, Maximum: 851660 usec, Average: 1376 usec, Peak to peak: 851659 usec, Stddev: 5447 usec, Sum: 258803984 usec
Based on the timestamps above, to calculate RTT, we need T1 and T4. In the case where SNMP-Agent sends the first 'get-request' to the RPM client between T1 to T4, RPM client has not received the T4. Since this is the first probe in the current test and there were no other probes sent earlier, we do not have any RTT for this current test untill we get the T4 for this probe. The 'pingResultsAverageRtt' will show as "0".
This behavior is working as designed for the current Junos releases.
JTAC recommends to use Juniper Networks enterprise-specific Ping MIB extends the standard Ping MIB control table (RFC 2925).
In this case, we can use jnxPingResultsRttUs (iso.3.6.1.4.1.2636.3.7.1.3.1.1) instead of pingResultsAverageRtt to monitor probe target reachability.
user@router> show snmp mib walk 1.3.6.1.4.1.2636.3.7.1.3.1.1 jnxPingResultsRttUs.3.50.49.57.3.50.49.57 = 795 jnxPingResultsRttUs.3.52.48.48.3.52.48.48 = 851 jnxPingResultsRttUs.3.52.48.51.3.52.48.51 = 780
For more information regarding jnxPingResultsRttUs, refer to: jnxPingResultsTable.