2016-04-03 09:15:55 +01:00
|
|
|
# prometheus
|
|
|
|
|
|
|
|
|
|
This module enables prometheus metrics for CoreDNS.
|
|
|
|
|
|
|
|
|
|
The following metrics are exported:
|
|
|
|
|
|
|
|
|
|
* coredns_dns_request_count_total
|
|
|
|
|
* coredns_dns_request_duration_seconds
|
|
|
|
|
* coredns_dns_response_size_bytes
|
|
|
|
|
* coredns_dns_response_rcode_count_total
|
|
|
|
|
|
|
|
|
|
Each counter has a label `zone` which is the zonename used for the request/response.
|
|
|
|
|
The `request_count` metrics has an extra label `qtype` which holds the qtype. And
|
|
|
|
|
`rcode_count` has an extra label which has the rcode.
|
|
|
|
|
|
|
|
|
|
Restarting CoreDNS will stop the monitoring. This is a bug. Also [this upstream
|
|
|
|
|
Caddy bug](https://github.com/mholt/caddy/issues/675).
|
|
|
|
|
|
|
|
|
|
## Syntax
|
|
|
|
|
|
|
|
|
|
~~~
|
|
|
|
|
prometheus
|
|
|
|
|
~~~
|
|
|
|
|
|
|
|
|
|
For each zone that you want to see metrics for.
|
|
|
|
|
|
|
|
|
|
It optionally takes an address where the metrics are exported, the default
|
2016-04-05 16:23:03 +01:00
|
|
|
is `localhost:9154`. The metrics path is fixed to `/metrics`.
|
2016-04-03 09:15:55 +01:00
|
|
|
|
|
|
|
|
## Examples
|
