check_mtr

Check_mtr is a Nagios compatible monitoring plugin written in Python. It compares values retrieved from mtr, like average latency and package loss, to the expected values given to the script by the CLI. If any value does not meet the expectations it returns CRITICAL (exit code 2).

There are three different values that the plugin is able to check. Every value has a different CLI option. If the option is given then the value will be checked, otherwise it will not be checked.

Usage

Option	Description	required?
-h	Help	❌
-H/--host	The host to run mtr on	✅
-j/--jumps	The routing regex that should match the routing path	❌
-l/--latency	The maximum expected latency (ms)	❌
-p/--packetloss	The maximum expected packet loss (%)	❌
-r/--routers	The routers on the routing path (Order is unimportant)	❌
-4/--ipv4	Use IPv4 to execute mtr	❌
-6/--ipv6	Use IPv6 to execute mtr	❌

At least one of the options -j, -l and -p must be set, because otherwise nothing will be checked. In this case the script returns UNKNOWN (exit code 3). By default mtr is executed with IPv4, if you want to execute it with IPv6 add the -6/--ipv6 flag to the CLI. If by accident both are given, then IPv4 will be used.

The latency and packet loss must be smaller than the maximum expected value for the check to be successful. The routing path is checked by looking at the routing regex and then decide whether the regex matches the actual routing path (returned by mtr) or not.

Example:

python3 check_mtr.py -H example.com -l 100 -p 5 -j "*1-5,12.13.14.15,*1,123.124.125.126[15ms:1%]" -r "[123.124.125.126, 12.13.14.15]"

Routing Regex

The routing regex can be build by the following rules:

1. routing := routing,routing | wildcard | <router address>values
2. wildcard := * | *<int> | *<int> - <int>
3. values := [latency:package loss] | Ɛ
4. latency := <int> # in ms
5. package loss := <int> # in %

A router address can be an IP or a hostname

If the latency and package loss are given after a hostname in the regex then these are the maximum values for latency and package loss that this host may have. If the actual values exceed these expected ones, the plugin will return CRITICAL.

Meaning of the wildcard cases:

1. * = Any number of unspecified router addresses may follow
2. *1 = One unspecified router address may follow
3. *1-3 = One to three unspecified router addresses may follow

You are also able to simply hardcode expected routes by only using router addresses concatenated with ","

Output + Performance Data

The plugin outputs the mtr results for the different hops on the routing path. The plugin also prints out the latency (rta) and package loss (pl) as performance data. So the values for latency and package loss will then be shown as performance data in your monitoring system after the check was executed. The performance data are always printed, also when the check fails. In the following there is an example of the output of the check plugin:

OK - All values were in the valid range
Hops:
1. 1.2.3.4 0.0% 10 3.234 15.401 5.308 59.167 10.101
2. 2.3.4.5 0.0% 10 1.783 2.097 1.365 5.015 10.73
3. 3.4.5.6 0.0% 10 4.485 4.611 4.367 5.143 0.219
4. 4.5.6.7 0.0% 10 49.007 9.351 4.832 49.007 13.933
5. 5.6.7.8 0.0% 10 8.936 8.999 8.766 9.104 0.111
6. 6.7.8.9 0.0% 10 9.425 9.981 9.425 10.98 0.482
7. 7.8.9.10 0.0% 10 8.89 8.88 8.797 8.925 0.041 | 'hop_1.2.3.4_rta'=10.401;; 'hop_1.2.3.4_pl'=0.0;; 'hop_2.3.4.5_rta'=1.097;; 'hop_2.3.4.5_pl'=0.0;; 'hop_3.4.5.6_rta'=4.611;; 'hop_3.4.5.6_pl'=0.0;; 'hop_4.5.6.7_rta'=9.351;; 'hop_4.5.6.7_pl'=0.0;; 'hop_5.6.7.8_rta'=8.999;; 'hop_5.6.7.8_pl'=0.0;; 'hop_6.7.8.9_rta'=9.981;; 'hop_6.7.8.9_pl'=0.0;; 'hop_7.8.9.10_rta'=8.88;; 'hop_7.8.9.10_pl'=0.0;;

Dependencies

The script has no dependencies, all imported libraries are available from the python standard library.