This package exports Audisto web crawler report data as Prometheus metrics.
Start the HTTP server like this:
$ AUDISTO_USERNAME=APIUSER AUDISTO_PASSWORD=APISECRET audisto_exporter --host=127.0.0.1 --port=9307
Pass --ttl=SECONDS
to cache API results for the given time or -1 to disable (default is 600).
Prometheus considers metrics stale after 300s, so that's the highest scrape_interval one should use.
However it's usually unnecessary to hit the API that often, since the information does not change that rapidly.
scrape_configs: - job_name: 'audisto' scrape_interval: 300s static_configs: - targets: ['localhost:9307']
We export two metrics, firstly a gauge http_requests_total
with labels {service="http://www.zeit.de/index",code="200"}
.
Status codes 601/602/603 are used for indexable/robots noindex/canonical.
Status codes 700/701/702/703 are used for duplicate title/description/body/similar urls.
Secondly we export a histogram response_time
,
with labels {service="http://www.zeit.de/index"}
and buckets as determined by Audisto.
Additionally, a audisto_scrape_duration_seconds
gauge is exported.