Using service RPM ping probe to detect link congestion

In Junos, we can use RPM (real-time performance monitoring) to monitor link quality.

1. give a name to the ping probe and test name

edit services rpm probe ping-probe test test-ping


2. set probe type

set probe-type icmp-ping


3. set probe count for each test and interval between each ping (in second)

set probe-count 10
set probe-interval 1

4. set source and target address (must be the interconnect IPs if you want to detect link quality)

set target address
set source address

5. set the thresold for sucessful probe

set thresolds total-loss 2
set std-dev-rtt 50000

where std-dev value is in micro second, not millisecond (1000 micro = 1 milli)


6. log the result in a remote monitoring (Syslog) server for generating alerts.

set system syslog host daemon any
set system syslog host match rmopd

where rmopd is the daemon name for ping probe