Version: v1.28.1 Latest Latest

This package is not in the latest version of its module.

Go to latest
Published: Sep 12, 2023 License: MIT Imports: 45 Imported by: 180


Prometheus Input Plugin

The prometheus input plugin gathers metrics from HTTP servers exposing metrics in Prometheus format.

Global configuration options

In addition to the plugin-specific configuration settings, plugins support additional global and plugin configuration settings. These settings are used to modify metrics, tags, and field or create aliases and configure ordering, etc. See the for more details.


# Read metrics from one or many prometheus clients
  ## An array of urls to scrape metrics from.
  urls = ["http://localhost:9100/metrics"]

  ## Metric version controls the mapping from Prometheus metrics into Telegraf metrics.
  ## See "Metric Format Configuration" in plugins/inputs/prometheus/ for details.
  ## Valid options: 1, 2
  # metric_version = 1

  ## Url tag name (tag containing scrapped url. optional, default is "url")
  # url_tag = "url"

  ## Whether the timestamp of the scraped metrics will be ignored.
  ## If set to true, the gather time will be used.
  # ignore_timestamp = false

  ## An array of Kubernetes services to scrape metrics from.
  # kubernetes_services = [""]

  ## Kubernetes config file to create client from.
  # kube_config = "/path/to/kubernetes.config"

  ## Scrape Pods
  ## Enable scraping of k8s pods. Further settings as to which pods to scape
  ## are determiend by the 'method' option below. When enabled, the default is
  ## to use annotations to determine whether to scrape or not.
  # monitor_kubernetes_pods = false

  ## Scrape Pods Method
  ## annotations: default, looks for specific pod annotations documented below
  ## settings: only look for pods matching the settings provided, not
  ##   annotations
  ## settings+annotations: looks at pods that match annotations using the user
  ##   defined settings
  # monitor_kubernetes_pods_method = "annotations"

  ## Scrape Pods 'annotations' method options
  ## If set method is set to 'annotations' or 'settings+annotations', these
  ## annotation flags are looked for:
  ## - Required to enable scraping for this pod. Can also
  ##     use '' annotation to opt-out entirely.
  ## - If the metrics endpoint is secured then you will
  ##     need to set this to 'https' & most likely set the tls config
  ## - If the metrics path is not /metrics, define it with
  ##     this annotation
  ## - If port is not 9102 use this annotation

  ## Scrape Pods 'settings' method options
  ## When using 'settings' or 'settings+annotations', the default values for
  ## annotations can be modified using with the following options:
  # monitor_kubernetes_pods_scheme = "http"
  # monitor_kubernetes_pods_port = "9102"
  # monitor_kubernetes_pods_path = "/metrics"

  ## Get the list of pods to scrape with either the scope of
  ## - cluster: the kubernetes watch api (default, no need to specify)
  ## - node: the local cadvisor api; for scalability. Note that the config node_ip or the environment variable NODE_IP must be set to the host IP.
  # pod_scrape_scope = "cluster"

  ## Only for node scrape scope: node IP of the node that telegraf is running on.
  ## Either this config or the environment variable NODE_IP must be set.
  # node_ip = ""

  ## Only for node scrape scope: interval in seconds for how often to get updated pod list for scraping.
  ## Default is 60 seconds.
  # pod_scrape_interval = 60

  ## Restricts Kubernetes monitoring to a single namespace
  ##   ex: monitor_kubernetes_pods_namespace = "default"
  # monitor_kubernetes_pods_namespace = ""
  ## The name of the label for the pod that is being scraped.
  ## Default is 'namespace' but this can conflict with metrics that have the label 'namespace'
  # pod_namespace_label_name = "namespace"
  # label selector to target pods which have the label
  # kubernetes_label_selector = "env=dev,app=nginx"
  # field selector to target pods
  # eg. To scrape pods on a specific node
  # kubernetes_field_selector = "spec.nodeName=$HOSTNAME"

  ## Filter which pod annotations and labels will be added to metric tags
  # pod_annotation_include = ["annotation-key-1"]
  # pod_annotation_exclude = ["exclude-me"]
  # pod_label_include = ["label-key-1"]
  # pod_label_exclude = ["exclude-me"]

  # cache refresh interval to set the interval for re-sync of pods list.
  # Default is 60 minutes.
  # cache_refresh_interval = 60

  ## Scrape Services available in Consul Catalog
  # [inputs.prometheus.consul]
  #   enabled = true
  #   agent = "http://localhost:8500"
  #   query_interval = "5m"

  #   [[inputs.prometheus.consul.query]]
  #     name = "a service name"
  #     tag = "a service tag"
  #     url = 'http://{{if ne .ServiceAddress ""}}{{.ServiceAddress}}{{else}}{{.Address}}{{end}}:{{.ServicePort}}/{{with .ServiceMeta.metrics_path}}{{.}}{{else}}metrics{{end}}'
  #     [inputs.prometheus.consul.query.tags]
  #       host = "{{.Node}}"

  ## Use bearer token for authorization. ('bearer_token' takes priority)
  # bearer_token = "/path/to/bearer/token"
  ## OR
  # bearer_token_string = "abc_123"

  ## HTTP Basic Authentication username and password. ('bearer_token' and
  ## 'bearer_token_string' take priority)
  # username = ""
  # password = ""

  ## Optional custom HTTP headers
  # http_headers = {"X-Special-Header" = "Special-Value"}

  ## Specify timeout duration for slower prometheus clients (default is 5s)
  # timeout = "5s"

  ## deprecated in 1.26; use the timeout option
  # response_timeout = "5s"

  ## HTTP Proxy support
  # use_system_proxy = false
  # http_proxy_url = ""

  ## Optional TLS Config
  # tls_ca = /path/to/cafile
  # tls_cert = /path/to/certfile
  # tls_key = /path/to/keyfile

  ## Use TLS but skip chain & host verification
  # insecure_skip_verify = false

  ## Use the given name as the SNI server name on each URL
  # tls_server_name = ""

  ## TLS renegotiation method, choose from "never", "once", "freely"
  # tls_renegotiation_method = "never"

  ## Enable/disable TLS
  ## Set to true/false to enforce TLS being enabled/disabled. If not set,
  ## enable TLS only if any of the other options are specified.
  # tls_enable = true

  ## Control pod scraping based on pod namespace annotations
  ## Pass and drop here act like tagpass and tagdrop, but instead
  ## of filtering metrics they filters pod candidates for scraping
  # annotation_key = ["value1", "value2"]
  # some_annotation_key = ["dont-scrape"]

urls can contain a unix socket as well. If a different path is required (default is /metrics for both http[s] and unix) for a unix socket, add path as a query parameter as follows: unix:///var/run/prometheus.sock?path=/custom/metrics

Metric Format Configuration

The metric_version setting controls how telegraf translates prometheus format metrics to telegraf metrics. There are two options.

With metric_version = 1, the prometheus metric name becomes the telegraf metric name. Prometheus labels become telegraf tags. Prometheus values become telegraf field values. The fields have generic keys based on the type of the prometheus metric. This option produces metrics that are dense (not sparse). Denseness is a useful property for some outputs, including those that are more efficient with row-oriented data.

metric_version = 2 differs in a few ways. The prometheus metric name becomes a telegraf field key. Metrics hold more than one value and the field keys aren't generic. The resulting metrics are sparse, but for some outputs they may be easier to process or query, including those that are more efficient with column-oriented data. The telegraf metric name is the same for all metrics in the input instance. It can be set with the name_override setting and defaults to "prometheus". To have multiple metric names, you can use multiple instances of the plugin, each with its own name_override.

metric_version = 2 uses the same histogram format as the histogram aggregator

The Example Outputs sections shows examples for both options.

When using this plugin along with the prometheus_client output, use the same option in both to ensure metrics are round-tripped without modification.

Kubernetes Service Discovery

URLs listed in the kubernetes_services parameter will be expanded by looking up all A records assigned to the hostname as described in Kubernetes DNS service discovery.

This method can be used to locate all Kubernetes headless services.

Kubernetes scraping

Enabling this option will allow the plugin to scrape for prometheus annotation on Kubernetes pods. Currently, you can run this plugin in your kubernetes cluster, or we use the kubeconfig file to determine where to monitor. Currently the following annotation are supported:

  • Enable scraping for this pod.
  • If the metrics endpoint is secured then you will need to set this to https & most likely set the tls config. (default 'http')
  • Override the path for the metrics endpoint on the service. (default '/metrics')
  • Used to override the port. (default 9102)

Using the monitor_kubernetes_pods_namespace option allows you to limit which pods you are scraping.

The setting pod_namespace_label_name allows you to change the label name for the namespace of the pod you are scraping. The default is namespace, but this will overwrite a label with the name namespace from a metric scraped.

Using pod_scrape_scope = "node" allows more scalable scraping for pods which will scrape pods only in the node that telegraf is running. It will fetch the pod list locally from the node's kubelet. This will require running Telegraf in every node of the cluster. Note that either node_ip must be specified in the config or the environment variable NODE_IP must be set to the host IP. ThisThe latter can be done in the yaml of the pod running telegraf:

  - name: NODE_IP
        fieldPath: status.hostIP

If using node level scrape scope, pod_scrape_interval specifies how often (in seconds) the pod list for scraping should updated. If not specified, the default is 60 seconds.

The pod running telegraf will need to have the proper rbac configuration in order to be allowed to call the k8s api to discover and watch pods in the cluster. A typical configuration will create a service account, a cluster role with the appropriate rules and a cluster role binding to tie the cluster role to the service account. Example of configuration for cluster level discovery:

kind: ClusterRole
  name: telegraf-k8s-role-{{.Release.Name}}
- apiGroups: [""]
  - nodes
  - nodes/proxy
  - services
  - endpoints
  - pods
  verbs: ["get", "list", "watch"]
# Rolebinding for namespace to cluster-admin
kind: ClusterRoleBinding
  name: telegraf-k8s-role-{{.Release.Name}}
  kind: ClusterRole
  name: telegraf-k8s-role-{{.Release.Name}}
- kind: ServiceAccount
  name: telegraf-k8s-{{ .Release.Name }}
  namespace: {{ .Release.Namespace }}
apiVersion: v1
kind: ServiceAccount
  name: telegraf-k8s-{{ .Release.Name }}
Consul Service Discovery

Enabling this option and configuring consul agent url will allow the plugin to query consul catalog for available services. Using query_interval the plugin will periodically query the consul catalog for services with name and tag and refresh the list of scraped urls. It can use the information from the catalog to build the scraped url and additional tags from a template.

Multiple consul queries can be configured, each for different service. The following example fields can be used in url or tag templates:

  • Node
  • Address
  • NodeMeta
  • ServicePort
  • ServiceAddress
  • ServiceTags
  • ServiceMeta

For full list of available fields and their type see struct CatalogService in

Bearer Token

If set, the file specified by the bearer_token parameter will be read on each interval and its contents will be appended to the Bearer string in the Authorization header.

Usage for Caddy HTTP server

Steps to monitor Caddy with Telegraf's Prometheus input plugin:

#   ## An array of urls to scrape metrics from.
  urls = ["http://localhost:2019/metrics"]

This is the default URL where Caddy will send data. For more details, please read the Caddy Prometheus documentation.


Measurement names are based on the Metric Family and tags are created for each label. The value is added to a field named based on the metric type.

All metrics receive the url tag indicating the related URL specified in the Telegraf configuration. If using Kubernetes service discovery the address tag is also added indicating the discovered ip address.

Example Output

# HELP go_gc_duration_seconds A summary of the GC invocation durations.
# TYPE go_gc_duration_seconds summary
go_gc_duration_seconds{quantile="0"} 7.4545e-05
go_gc_duration_seconds{quantile="0.25"} 7.6999e-05
go_gc_duration_seconds{quantile="0.5"} 0.000277935
go_gc_duration_seconds{quantile="0.75"} 0.000706591
go_gc_duration_seconds{quantile="1"} 0.000706591
go_gc_duration_seconds_sum 0.00113607
go_gc_duration_seconds_count 4
# HELP go_goroutines Number of goroutines that currently exist.
# TYPE go_goroutines gauge
go_goroutines 15
# HELP cpu_usage_user Telegraf collected metric
# TYPE cpu_usage_user gauge
cpu_usage_user{cpu="cpu0"} 1.4112903225816156
cpu_usage_user{cpu="cpu1"} 0.702106318955865
cpu_usage_user{cpu="cpu2"} 2.0161290322588776
cpu_usage_user{cpu="cpu3"} 1.5045135406226022
go_gc_duration_seconds,url= 1=0.001336611,count=14,sum=0.004527551,0=0.000057965,0.25=0.000083812,0.5=0.000286537,0.75=0.000365303 1505776733000000000
go_goroutines,url= gauge=21 1505776695000000000
cpu_usage_user,cpu=cpu0,url= gauge=1.513622603430151 1505776751000000000
cpu_usage_user,cpu=cpu1,url= gauge=5.829145728641773 1505776751000000000
cpu_usage_user,cpu=cpu2,url= gauge=2.119071644805144 1505776751000000000
cpu_usage_user,cpu=cpu3,url= gauge=1.5228426395944945 1505776751000000000
Output (when metric_version = 2)
prometheus,quantile=1,url= go_gc_duration_seconds=0.005574303 1556075100000000000
prometheus,quantile=0.75,url= go_gc_duration_seconds=0.0001046 1556075100000000000
prometheus,quantile=0.5,url= go_gc_duration_seconds=0.0000719 1556075100000000000
prometheus,quantile=0.25,url= go_gc_duration_seconds=0.0000579 1556075100000000000
prometheus,quantile=0,url= go_gc_duration_seconds=0.0000349 1556075100000000000
prometheus,url= go_gc_duration_seconds_count=324,go_gc_duration_seconds_sum=0.091340353 1556075100000000000
prometheus,url= go_goroutines=15 1556075100000000000
prometheus,cpu=cpu0,url= cpu_usage_user=1.513622603430151 1505776751000000000
prometheus,cpu=cpu1,url= cpu_usage_user=5.829145728641773 1505776751000000000
prometheus,cpu=cpu2,url= cpu_usage_user=2.119071644805144 1505776751000000000
prometheus,cpu=cpu3,url= cpu_usage_user=1.5228426395944945 1505776751000000000
Output with timestamp included

Below is an example of a Prometheus metric which includes a timestamp:

# TYPE test_counter counter
test_counter{label="test"} 1 1685443805885

Telegraf will generate the following metric:

test_counter,address=,label=test counter=1 1685443805885000000

using the standard configuration

  ## An array of urls to scrape metrics from.
  urls = ["http://localhost:2019/metrics"]

Please note: Metrics generated by Prometheus endpoints are generated with millisecond precision. The default Telegraf agent level precision setting reduces this to seconds. Change the precision setting at agent or plugin level to milliseconds or smaller to report metric timestamps with full precision.




This section is empty.


This section is empty.


func Parse added in v1.14.0

func Parse(buf []byte, header http.Header, ignoreTimestamp bool) ([]telegraf.Metric, error)


type ConsulConfig added in v1.20.0

type ConsulConfig struct {
	// Address of the Consul agent. The address must contain a hostname or an IP address
	// and optionally a port (format: "host:port").
	Enabled       bool            `toml:"enabled"`
	Agent         string          `toml:"agent"`
	QueryInterval config.Duration `toml:"query_interval"`
	Queries       []*ConsulQuery  `toml:"query"`

type ConsulQuery added in v1.20.0

type ConsulQuery struct {
	// A name of the searched services (not ID)
	ServiceName string `toml:"name"`

	// A tag of the searched services
	ServiceTag string `toml:"tag"`

	// A DC of the searched services
	ServiceDc string `toml:"dc"`

	// A template URL of the Prometheus gathering interface. The hostname part
	// of the URL will be replaced by discovered address and port.
	ServiceURL string `toml:"url"`

	// Extra tags to add to metrics found in Consul
	ServiceExtraTags map[string]string `toml:"tags"`
	// contains filtered or unexported fields

One Consul service discovery query

type MonitorMethod added in v1.25.0

type MonitorMethod string
const (
	MonitorMethodNone                   MonitorMethod = ""
	MonitorMethodAnnotations            MonitorMethod = "annotations"
	MonitorMethodSettings               MonitorMethod = "settings"
	MonitorMethodSettingsAndAnnotations MonitorMethod = "settings+annotations"

type PodID added in v1.25.1

type PodID string

type Prometheus

type Prometheus struct {
	// An array of urls to scrape metrics from.
	URLs []string `toml:"urls"`

	// An array of Kubernetes services to scrape metrics from.
	KubernetesServices []string

	// Location of kubernetes config file
	KubeConfig string

	// Label Selector/s for Kubernetes
	KubernetesLabelSelector string `toml:"kubernetes_label_selector"`

	// Field Selector/s for Kubernetes
	KubernetesFieldSelector string `toml:"kubernetes_field_selector"`

	// Consul SD configuration
	ConsulConfig ConsulConfig `toml:"consul"`

	// Bearer Token authorization file path
	BearerToken       string `toml:"bearer_token"`
	BearerTokenString string `toml:"bearer_token_string"`

	// Basic authentication credentials
	Username string `toml:"username"`
	Password string `toml:"password"`

	HTTPHeaders map[string]string `toml:"http_headers"`

	ResponseTimeout config.Duration `toml:"response_timeout" deprecated:"1.26.0;use 'timeout' instead"`

	MetricVersion int `toml:"metric_version"`

	URLTag string `toml:"url_tag"`

	IgnoreTimestamp bool `toml:"ignore_timestamp"`

	Log telegraf.Logger


	// Should we scrape Kubernetes services for prometheus annotations
	MonitorPods           bool   `toml:"monitor_kubernetes_pods"`
	PodScrapeScope        string `toml:"pod_scrape_scope"`
	NodeIP                string `toml:"node_ip"`
	PodScrapeInterval     int    `toml:"pod_scrape_interval"`
	PodNamespace          string `toml:"monitor_kubernetes_pods_namespace"`
	PodNamespaceLabelName string `toml:"pod_namespace_label_name"`

	MonitorKubernetesPodsMethod MonitorMethod `toml:"monitor_kubernetes_pods_method"`
	MonitorKubernetesPodsScheme string        `toml:"monitor_kubernetes_pods_scheme"`
	MonitorKubernetesPodsPath   string        `toml:"monitor_kubernetes_pods_path"`
	MonitorKubernetesPodsPort   int           `toml:"monitor_kubernetes_pods_port"`

	NamespaceAnnotationPass map[string][]string `toml:"namespace_annotation_pass"`
	NamespaceAnnotationDrop map[string][]string `toml:"namespace_annotation_drop"`

	PodAnnotationInclude []string `toml:"pod_annotation_include"`
	PodAnnotationExclude []string `toml:"pod_annotation_exclude"`

	PodLabelInclude []string `toml:"pod_label_include"`
	PodLabelExclude []string `toml:"pod_label_exclude"`

	// Only for monitor_kubernetes_pods=true
	CacheRefreshInterval int `toml:"cache_refresh_interval"`
	// contains filtered or unexported fields

func (*Prometheus) AddressToURL added in v1.14.0

func (p *Prometheus) AddressToURL(u *url.URL, address string) *url.URL

func (*Prometheus) Gather

func (p *Prometheus) Gather(acc telegraf.Accumulator) error

Reads stats from all configured servers accumulates stats. Returns one of the errors encountered while gather stats (if any).

func (*Prometheus) GetAllURLs added in v1.14.0

func (p *Prometheus) GetAllURLs() (map[string]URLAndAddress, error)

func (*Prometheus) Init added in v1.14.0

func (p *Prometheus) Init() error

func (*Prometheus) SampleConfig

func (*Prometheus) SampleConfig() string

func (*Prometheus) Start added in v1.14.0

func (p *Prometheus) Start(_ telegraf.Accumulator) error

Start will start the Kubernetes and/or Consul scraping if enabled in the configuration

func (*Prometheus) Stop added in v1.14.0

func (p *Prometheus) Stop()

type URLAndAddress added in v1.14.0

type URLAndAddress struct {
	OriginalURL *url.URL
	URL         *url.URL
	Address     string
	Tags        map[string]string
	Namespace   string

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL