DNSmetrics connects to your accounts at multiple managed DNS providers using their APIs and emits standardized metrics in statsd format for easy ingestion into your monitoring solution of choice.
While any DNS provider has a control panel, their UI can be incomplete and they might not have alerting on metrics that are important to you. Furthermore, if you use several providers for redundancy, there is no way to have one dashboard that shows unified health of your DNS architecture.
Supported DNS providers
It is easy to add new providers. Contributions are most welcome.
Installation
With Go 1.7 installed, do make install
and this will result in a binary.
Configuration
Take a look at config.yml
. It should be self explanatory.
Running DNSmetrics
For testing, using --once
will output the collected metrics to stdout
instead of sending them to statsd. Without --once
, DNSmetrics will run
as a service, collecting and sending metrics every check_interval
.
List of Metrics
All metrics are tagged with zone
and provider
tags. Zone is the DNS
zone for the metric and provider is the DNS provider.
All metrics have the prefix dnsmetrics
.
Not all providers support all metrics described below.
-
zone.type.primary
,zone.type.secondary
- 0 or 1. -
zone.qps
- current (or most recent) rate of queries per second, of all types, for the zone. -
zone.record_count
- number of records of all types in the zone. -
zone.serial
- the serial number of the zone. -
zone.secondary.is_ok
- 0 or 1. 1 when provider considers the zone to be healthy. -
zone.secondary.is_expired
- 0 or 1. 1 when the zone is expired and presumably no longer being served by the provider. -
zone.secondary.sec_since_last_xfr
- time since last zone transfer.
Monitoring
A DataDog dashboard specification with some of the collected metrics can
be found in monitoring/datadog
. It can be imported
using the API.
Contributing
Fork it, create a new feature branch, make your changes, open a pull request. Tests would be much appreciated.