Do not track HTTP endpoints

[ochaplashkin]

The service has two endpoints:

1. GET / - list of events;
2. POST / - webhook handler.

We have added standard metrics in the #27. So there are scanning systems that are trying to find a vulnerability in our service. Now, the metrics for incoming requests look too big:

$ curl https://docbot.tarantool.io/metrics > metrics.txt
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 12.6M  100 12.6M    0     0  1061k      0  0:00:12  0:00:12 --:--:-- 1715k

Size of metrics is 12.6M (!)

Prometheus takes such a volume every 10 seconds, which leads to an instant expansion of timeseries storage and the inactivity of Prometheus.