atlas_exporter

Metric exporter for RIPE Atlas measurement results

Remarks

• this is an early version, more features will be added step by step
• at the moment only the last result of an measurement is used
• the required Go version is 1.11+

Streaming API

Since version 0.8 atlas_exporter also supports retrieving measurement results by RIPE Atlas Streaming API (https://atlas.ripe.net/docs/result-streaming/). Using this feature requires config file mode. All configured measurements are subscribed on start so the latest result for each probe is updated continuously and scrape time is reduced significantly. When a socket.io connection fails or times out a reconnect is initiated. The timeout can be configured using the -streaming.timeout parameter. Streaming API is the default for config file mode, it can be disabled by setting -streaming to false.

Histograms

Since version 1.0 atlas_exporter provides you with histograms of round trip times of the following measurement types:

• DNS
• Ping
• Traceroute
• HTTP

The buckets can be configured in the config file (see below).

Since this feature relies strongly on getting each update for a measurement, the Stream API mode has to be used. Histogram metrics enables you to calculate percentiles for a specifiv indicator (in our case round trip time). This allows better monitoring of defined service level objectives (e.g. Ping RTT of a specific measurement should be under a specific threshold based on 90% of the requests disregarding the highest 10% -> p90).

For more information: https://prometheus.io/docs/practices/histograms/

Install

go get -u github.com/czerwonk/atlas_exporter

Docker

To start the server:

docker run -d --restart unless-stopped -p 9400:9400 czerwonk/atlas_exporter

To run in config file mode:

docker run -d -e CONFIG=/tmp/config.yml -v /tmp/config.yml:/tmp/config.yml --restart unless-stopped -p 9400:9400 czerwonk/atlas_exporter 

Usage

Start server

./atlas_exporter

or using config file mode:

./atlas_exporter -config.file config.yml

Config file

for this example we want to retrieve results for measurement 8772164

measurements:
  - id: 8772164
    timeout: 120s
histogram_buckets:
  ping:
    rtt:
      - 5.0
      - 10.0
      - 25.0
      - 50.0
      - 100.0
filter_invalid_results: true

Call metrics URI

when using config file mode:

curl http://127.0.0.1:9400/metrics

or ad hoc for measuremnt 8772164:

curl http://127.0.0.1:9400/metrics?measurement_id=8772164

in both cases the result should look similar to this one:

# HELP atlas_traceroute_hops Number of hops
# TYPE atlas_traceroute_hops gauge
atlas_traceroute_hops{asn="1101",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="6031"} 9
atlas_traceroute_hops{asn="11051",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="17833"} 8
atlas_traceroute_hops{asn="111",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="6231"} 9
atlas_traceroute_hops{asn="11427",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="1121"} 13
atlas_traceroute_hops{asn="12337",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="267"} 13
atlas_traceroute_hops{asn="1257",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="140"} 11
atlas_traceroute_hops{asn="12586",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="2088"} 13
atlas_traceroute_hops{asn="12597",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="2619"} 10
atlas_traceroute_hops{asn="12714",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="2684"} 9
atlas_traceroute_hops{asn="133752",dst_addr="8.8.8.8",dst_name="8.8.8.8",ip_version="4",measurement="8772164",probe="6191"} 14

...

Features

• ping measurements (success, min/max/avg latency, dups, size)
• traceroute measurements (success, hop count, rtt)
• ntp (delay, derivation, ntp version)
• dns (succress, rtt)
• http (return code, rtt, http version, header size, body size)
• sslcert (alert, rtt)

Prometheus configuration

Ad-Hoc Mode

  - job_name: 'atlas_exporter'
    scrape_interval: 5m
    static_configs:
      - targets:
        - 7924888
        - 7924886
    relabel_configs:
      - source_labels: [__address__]
        regex: (.*)(:80)?
        target_label: __param_measurement_id
        replacement: ${1}
      - source_labels: [__param_measurement_id]
        regex: (.*)
        target_label: instance
        replacement: ${1}
      - source_labels: []
        regex: .*
        target_label: __address__
        replacement: atlas-exporter.mytld:9400

Config Mode

  - job_name: 'atlas_exporter'
    scrape_interval: 5m
    static_configs:
      - targets:
          - atlas-exporter.mytld:9400

Third Party Components

This software uses components of the following projects

• Go bindings for RIPE Atlas API (https://github.com/DNS-OARC/ripeatlas)
• Prometheus Go client library (https://github.com/prometheus/client_golang)

License

(c) Daniel Czerwonk, 2017. Licensed under LGPL3 license.

Prometheus

see https://prometheus.io/

The RIPE Atlas Project

see http://atlas.ripe.net

Further reading

I wrote an article about atlas_exporter for RIPE Labs. It covers version 0.5. https://labs.ripe.net/Members/daniel_czerwonk/using-ripe-atlas-measurement-results-in-prometheus-with-atlas_exporter