Junos SRX RPM

RPM { RealTime Probe Monitor } is  Juniper equal to the cisco IPSLA. It's supported in  various Junos devices. In this post we will look at the   icmp RTM that was setup for gathering uplink statistics

[edit]

kfelix@MANHATTAN# show | display set | match rpm                               

set services rpm probe ICMP test icmp1 probe-type icmp-ping

set services rpm probe ICMP test icmp1 target address 8.8.4.4

set services rpm probe ICMP test icmp1 probe-count 1

set services rpm probe ICMP test icmp1 probe-interval 120

set services rpm probe ICMP test icmp1 test-interval 120

[edit]

And dump of any active probe;

 "show services rpm probe-results owner ICMP test icmp1"

    Owner: ICMP, Test: icmp1

    Target address: 8.8.4.4, Probe type: icmp-ping, Test size: 1 probes

    Probe results:

      Response received

      Sun Nov 18 13:12:55 2018

      Sun Nov 18 13:12:55 2018, No hardware timestamps

      Rtt: 9893 usec, Round trip jitter: 140 usec, Round trip interarrival jitter: 4996 usec

    Results over current test:

      Probes sent: 1, Probes received: 1, Loss percentage: 0.000000

      Measurement: Round trip time

        Samples: 1, Minimum: 9893 usec, Maximum: 9893 usec, Average: 9893 usec, Peak to peak: 0 usec, Stddev: 0 usec, Sum: 9893 usec

      Measurement: Positive round trip jitter

        Samples: 1, Minimum: 140 usec, Maximum: 140 usec, Average: 140 usec, Peak to peak: 0 usec, Stddev: 0 usec, Sum: 140 usec

    Results over last test:

      Probes sent: 1, Probes received: 1, Loss percentage: 0.000000

      Test completed on Sun Nov 18 13:12:55 2018

      Measurement: Round trip time

        Samples: 1, Minimum: 9893 usec, Maximum: 9893 usec, Average: 9893 usec, Peak to peak: 0 usec, Stddev: 0 usec, Sum: 9893 usec

      Measurement: Positive round trip jitter

        Samples: 1, Minimum: 140 usec, Maximum: 140 usec, Average: 140 usec, Peak to peak: 0 usec, Stddev: 0 usec, Sum: 140 usec

    Results over all tests:

      Probes sent: 3995, Probes received: 3980, Loss percentage: 0.375469

      Measurement: Round trip time

        Samples: 3980, Minimum: 9442 usec, Maximum: 134912 usec, Average: 11707 usec, Peak to peak: 125470 usec, Stddev: 7648 usec, Sum: 46592163 usec

      Measurement: Positive round trip jitter

        Samples: 1994, Minimum: 0 usec, Maximum: 125086 usec, Average: 3567 usec, Peak to peak: 125086 usec, Stddev: 10224 usec, Sum: 7112597 usec

      Measurement: Negative round trip jitter

        Samples: 1985, Minimum: 1 usec, Maximum: 125069 usec, Average: 3583 usec, Peak to peak: 125068 usec, Stddev: 10229 usec, Sum: 7112336 usec

So the  last lines would be our collective count for all test and for the number of probes sent. I bold out a few lines of interest.

Keep in mind that various probe types exists and your not limited to just icmp types.

Possible completions:

  http-get             Perform HTTP Get request at target URL

  http-metadata-get    Perform HTTP Get request of metadata at target URL

  icmp-ping            Send ICMP echo request to target address

  icmp-ping-timestamp  Send ICMP timestamp request to target address

  icmp6-ping           Send ICMP6 echo request to target address

  tcp-ping             Send TCP packets to target

  udp-ping             Send UDP packets to target

  udp-ping-timestamp   Send UDP packets with timestamp to target

[edit]

These problems can be called out and used to help ensure and measure service availability. If you define  thresholds, you can  set traps on fails.

NSE ( network security expert) and Route/Switching Engineer

kfelix  -----a----t---- socpuppets ---dot---com

     ^      ^

=(  @  @ )=

         o

        /  \