README
¶
dogstatsd
Name
dogstatsd - publish coredns metrics to dogstatsd agents
Description
CoreDNS has a builtin mechanism for exposing metrics to Prometheus over a HTTP endpoint, but infrastructures that use a metric collection system based on dogstatsd have to run a bridge to poll CoreDNS's metrics endpoint and push them to a dogstatsd agent.
The dogstatsd plugin removes the need for a such bridge by allowing CoreDNS to directly push its metrics to a dogstatsd agent over UDP.
Syntax
dogstatsd [ADDR:PORT]
- ADDR Address at which a dogstatsd agent is available. It may be prefixed with udp://, udp4://, udp6://, or unixgram:// to indicate the protocal to use to push metrics to a dogstatsd agent. If unxigram:// is specified the address must be a path to a unix domain socket on the file system.
- PORT Port number at which the dogstatsd agent is accepting metrics. The port must not be set when the unixgram:// protocol is used to push metrics to a dogstatsd agent.
If you want more control:
dogstatsd [ADDR:PORT] {
buffer SIZE
flush INTERVAL
go
process
}
- buffer configures the size of the client buffer used to push metrics to a dogstatsd agent. This must not exceed the size of the receive buffer used by the agent. The minimum size is 512 B, the maximum is 64 KB.
- flush configures the time interval between flushes of metrics to a dogstatsd agent. The minimum interval is 1 second, there is not maximum.
- go enables reporting of go metrics to the dogstatsd agent.
- process enables reporting of process metrics to the dogstatsd agent.
Examples
Enable the dogstatsd plugin with a client buffer size of 8 KB, and flushing metrics every 10 seconds.
. {
dogstatsd localhost:8125 {
buffer 8192
flush 10s
}
}
plugins.cfg
This plugin is intended to appear right after the prometheus plugin declaration.
Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Dogstatsd ¶
type Dogstatsd struct {
Next plugin.Handler
// Address of the dogstatsd agent to push metrics to.
Addr string
// Size of the socket buffer used to push metrics to the dogstatsd agent.
BufferSize int
// Time interval between flushes of metrics to the dogstasd agent.
FlushInterval time.Duration
// Reg is the prometheus registry used by the metrics plugin where all
// metrics are registered.
Reg *prometheus.Registry
// Those flags control whether this plugin instance is allowed to report
// go and process metrics. This is used because those metrics are global
// so it is OK if only a single plugin pushes them to the dogstatsd agent.
EnableGoMetrics bool
EnableProcessMetrics bool
// ZoneNames is the list of zones that this plugin reports metrics for.
ZoneNames []string
// contains filtered or unexported fields
}
Dogstatsd is the implementation of the dogstasd coredns plugin.
Source Files