Important: This documentation is about an older version. It's relevant only to the release noted, many of the features and functions have been updated or replaced. Please view the current version.

Send data

Promtail

Configure service discovery

Open source

Configuring Promtail for service discovery

File Target Discovery

Promtail discovers locations of log files and extract labels from them through the scrape_configs section in the config YAML. The syntax is identical to what Prometheus uses.

scrape_configs contains one or more entries which are executed for each discovered target (i.e., each container in each new pod running in the instance):

scrape_configs:
  - job_name: local
    static_configs:
      - ...

  - job_name: kubernetes
    kubernetes_sd_config:
      - ...

If more than one scrape config section matches your logs, you will get duplicate entries as the logs are sent in different streams likely with slightly different labels.

There are different types of labels present in Promtail:

Labels starting with __ (two underscores) are internal labels. They usually come from dynamic sources like service discovery. Once relabeling is done, they are removed from the label set. To persist internal labels so they’re sent to Grafana Loki, rename them so they don’t start with __. See Relabeling for more information.
Labels starting with __meta_kubernetes_pod_label_* are “meta labels” which are generated based on your Kubernetes pod’s labels.
For example, if your Kubernetes pod has a label name set to foobar, then the scrape_configs section will receive an internal label __meta_kubernetes_pod_label_name with a value set to foobar.
Other labels starting with __meta_kubernetes_* exist based on other Kubernetes metadata, such as the namespace of the pod (__meta_kubernetes_namespace) or the name of the container inside the pod (__meta_kubernetes_pod_container_name). Refer to the Prometheus docs for the full list of Kubernetes meta labels.
The __path__ label is a special label which Promtail uses after discovery to figure out where the file to read is located. Wildcards are allowed, for example /var/log/*.log to get all files with a log extension in the specified directory, and /var/log/**/*.log for matching files and directories recursively. For a full list of options check out the docs for the library Promtail uses.
The __path_exclude__ label is another special label Promtail uses after discovery, to exclude a subset of the files discovered using __path__ from being read in the current scrape_config block. It uses the same library to enable usage of wildcards and glob patterns.
The label filename is added for every file found in __path__ to ensure the uniqueness of the streams. It is set to the absolute path of the file the line was read from.

Kubernetes Discovery

While Promtail can use the Kubernetes API to discover pods as targets, it can only read log files from pods that are running on the same node as the one Promtail is running on. Promtail looks for a __host__ label on each target and validates that it is set to the same hostname as Promtail’s (using either $HOSTNAME or the hostname reported by the kernel if the environment variable is not set).

This means that any time Kubernetes service discovery is used, there must be a relabel_config that creates the intermediate label __host__ from __meta_kubernetes_pod_node_name:

relabel_configs:
  - source_labels: ['__meta_kubernetes_pod_node_name']
    target_label: '__host__'

See Relabeling for more information. For more information on how to configure the service discovery see the Kubernetes Service Discovery configuration.

Journal Scraping (Linux Only)

On systems with systemd, Promtail also supports reading from the journal. Unlike file scraping which is defined in the static_configs stanza, journal scraping is defined in a journal stanza:

scrape_configs:
  - job_name: journal
    journal:
      json: false
      max_age: 12h
      path: /var/log/journal
      matches: _TRANSPORT=kernel
      labels:
        job: systemd-journal
    relabel_configs:
      - source_labels: ['__journal__systemd_unit']
        target_label: 'unit'

All fields defined in the journal section are optional, and are just provided here for reference. The max_age field ensures that no older entry than the time specified will be sent to Loki; this circumvents “entry too old” errors. The path field tells Promtail where to read journal entries from. The labels map defines a constant list of labels to add to every journal entry that Promtail reads. The matches field adds journal filters. If multiple filters are specified matching different fields, the log entries are filtered by both, if two filters apply to the same field, then they are automatically matched as alternatives.

When the json field is set to true, messages from the journal will be passed through the pipeline as JSON, keeping all of the original fields from the journal entry. This is useful when you don’t want to index some fields but you still want to know what values they contained.

By default, Promtail reads from the journal by looking in the /var/log/journal and /run/log/journal paths. If running Promtail inside of a Docker container, the path appropriate to your distribution should be bind mounted inside of Promtail along with binding /etc/machine-id. Bind mounting /etc/machine-id to the path of the same name is required for the journal reader to know which specific journal to read from. For example:

docker run \
  -v /var/log/journal/:/var/log/journal/ \
  -v /run/log/journal/:/run/log/journal/ \
  -v /etc/machine-id:/etc/machine-id \
  grafana/promtail:latest \
  -config.file=/path/to/config/file.yaml

When Promtail reads from the journal, it brings in all fields prefixed with __journal_ as internal labels. Like in the example above, the _SYSTEMD_UNIT field from the journal was transformed into a label called unit through relabel_configs. See Relabeling for more information, also look at the systemd man pages for a list of fields exposed by the journal.

Here’s an example where the SYSTEMD_UNIT, HOSTNAME, and SYSLOG_IDENTIFIER are relabeled for use in Loki.

Keep in mind that labels prefixed with __ will be dropped, so relabeling is required to keep these labels.

- job_name: systemd-journal
  journal:
    labels:
      cluster: ops-tools1
      job: default/systemd-journal
    path: /var/log/journal
  relabel_configs:
  - source_labels:
    - __journal__systemd_unit
    target_label: systemd_unit
  - source_labels:
    - __journal__hostname
    target_label: nodename
  - source_labels:
    - __journal_syslog_identifier
    target_label: syslog_identifier

Windows Event Log

On Windows Promtail supports reading from the event log. Windows event targets can be configured using the windows_events stanza:

scrape_configs:
- job_name: windows
  windows_events:
    use_incoming_timestamp: false
    bookmark_path: "./bookmark.xml"
    eventlog_name: "Application"
    xpath_query: '*'
    labels:
      job: windows
  relabel_configs:
    - source_labels: ['computer']
      target_label: 'host'

When Promtail receives an event it will attach the channel and computer labels and serialize the event in json. You can relabel default labels via Relabeling if required.

Providing a path to a bookmark is mandatory, it will be used to persist the last event processed and allow resuming the target without skipping logs.

Read the configuration section for more information.

See the eventlogmessage stage for extracting data from the message.

GCP Log scraping

Promtail supports scraping cloud resource logs such as GCS bucket logs, load balancer logs, and Kubernetes cluster logs from GCP. Configuration is specified in the gcplog section, within scrape_config.

There are two kind of scraping strategies: pull and push.

Pull

  - job_name: gcplog
    gcplog:
      subscription_type: "pull" # If the `subscription_type` field is empty, defaults to `pull`
      project_id: "my-gcp-project"
      subscription: "my-pubsub-subscription"
      use_incoming_timestamp: false # default rewrite timestamps.
      use_full_line: false # default use textPayload as log line.
      labels:
        job: "gcplog"
    relabel_configs:
      - source_labels: ['__gcp_resource_type']
        target_label: 'resource_type'
      - source_labels: ['__gcp_resource_labels_project_id']
        target_label: 'project'

Here project_id and subscription are the only required fields.

project_id is the GCP project id.
subscription is the GCP pubsub subscription where Promtail can consume log entries from.

Before using gcplog target, GCP should be configured with pubsub subscription to receive logs from.

It also supports relabeling and pipeline stages just like other targets.

When Promtail receives GCP logs, various internal labels are made available for relabeling:

__gcp_logname
__gcp_severity
__gcp_resource_type
__gcp_resource_labels_<NAME>
__gcp_labels_<NAME> In the example above, the project_id label from a GCP resource was transformed into a label called project through relabel_configs.

Push

  - job_name: gcplog
    gcplog:
      subscription_type: "push"
      use_incoming_timestamp: false
      labels:
        job: "gcplog-push"
      server:
        http_listen_address: 0.0.0.0
        http_listen_port: 8080
    relabel_configs:
      - source_labels: ['__gcp_message_id']
        target_label: 'message_id'
      - source_labels: ['__gcp_attributes_logging_googleapis_com_timestamp']
        target_label: 'incoming_ts'

When configuring the GCP Log push target, Promtail will start an HTTP server listening on port 8080, as configured in the server section. This server exposes the single endpoint POST /gcp/api/v1/push, responsible for receiving logs from GCP.

For Google’s PubSub to be able to send logs, Promtail server must be publicly accessible, and support HTTPS. For that, Promtail can be deployed as part of a larger orchestration service like Kubernetes, which can handle HTTPS traffic through an ingress, or it can be hosted behind a proxy/gateway, offloading the HTTPS to that component and routing the request to Promtail. Once that’s solved, GCP can be configured to send logs to Promtail.

It also supports relabeling and pipeline stages.

When Promtail receives GCP logs, various internal labels are made available for relabeling:

__gcp_message_id
__gcp_subscription_name
__gcp_attributes_<NAME>
__gcp_logname
__gcp_severity
__gcp_resource_type
__gcp_resource_labels_<NAME>
__gcp_labels_<NAME>

In the example above, the __gcp_message_id and the __gcp_attributes_logging_googleapis_com_timestamp labels are transformed to message_id and incoming_ts through relabel_configs. All other internal labels, for example some other attribute, will be dropped by the target if not transformed.

Syslog Receiver

Promtail supports receiving IETF Syslog (RFC5424) messages from a TCP or UDP stream. Receiving syslog messages is defined in a syslog stanza:

scrape_configs:
  - job_name: syslog
    syslog:
      listen_address: 0.0.0.0:1514
      listen_protocol: tcp
      idle_timeout: 60s
      label_structured_data: yes
      labels:
        job: "syslog"
    relabel_configs:
      - source_labels: ['__syslog_message_hostname']
        target_label: 'host'

The only required field in the syslog section is the listen_address field, where a valid network address must be provided. The default protocol for receiving messages is TCP. To change the protocol, the listen_protocol field can be changed to udp. Note, that UDP does not support TLS. The idle_timeout can help with cleaning up stale syslog connections. If label_structured_data is set, structured data in the syslog header will be translated to internal labels in the form of __syslog_message_sd_<ID>_<KEY>. The labels map defines a constant list of labels to add to every journal entry that Promtail reads.

Note that it is recommended to deploy a dedicated syslog forwarder like syslog-ng or rsyslog in front of Promtail. The forwarder can take care of the various specifications and transports that exist (UDP, BSD syslog, …). See recommended output configurations for syslog-ng and rsyslog.

When Promtail receives syslog messages, it brings in all header fields, parsed from the received message, prefixed with __syslog_ as internal labels. Like in the example above, the __syslog_message_hostname field from the journal was transformed into a label called host through relabel_configs. See Relabeling for more information.

Syslog-NG Output Configuration

destination d_loki {
  syslog("localhost" transport("tcp") port(<promtail_port>));
};

Rsyslog Output Configuration

For sending messages via TCP:

*.* action(type="omfwd" protocol="tcp" target="<promtail_host>" port="<promtail_port>" Template="RSYSLOG_SyslogProtocol23Format" TCP_Framing="octet-counted" KeepAlive="on")

For sending messages via UDP:

*.* action(type="omfwd" protocol="udp" target="<promtail_host>" port="<promtail_port>" Template="RSYSLOG_SyslogProtocol23Format")

Azure Event Hubs

Promtail supports reading messages from Azure Event Hubs. Targets can be configured using the azure_event_hubs stanza:

- job_name: azure_event_hubs
  azure_event_hubs:
    group_id: "mygroup"
    fully_qualified_namespace: my-namespace.servicebus.windows.net:9093
    connection_string: "my-connection-string"
    event_hubs:
      - event-hub-name
    labels:
      job: azure_event_hub
  relabel_configs:
    - action: replace
      source_labels:
        - __azure_event_hubs_category
      target_label: category

Only fully_qualified_namespace, connection_string and event_hubs are required fields. Read the configuration section for more information.

Kafka

Promtail supports reading message from Kafka using a consumer group. The Kafka targets can be configured using the kafka stanza:

scrape_configs:
- job_name: kafka
  kafka:
    brokers:
    - my-kafka-0.org:50705
    - my-kafka-1.org:50705
    topics:
    - ^promtail.*
    - some_fixed_topic
    labels:
      job: kafka
  relabel_configs:
      - action: replace
        source_labels:
          - __meta_kafka_topic
        target_label: topic
      - action: replace
        source_labels:
          - __meta_kafka_partition
        target_label: partition
      - action: replace
        source_labels:
          - __meta_kafka_group_id
        target_label: group
      - action: replace
        source_labels:
          - __meta_kafka_message_key
        target_label: message_key

Only the brokers and topics are required. Read the configuration section for more information.

GELF

GELF support in Promtail is an experimental feature.

Promtail supports listening message using the GELF UDP protocol. The GELF targets can be configured using the gelf stanza:

scrape_configs:
- job_name: gelf
  gelf:
    listen_address: "0.0.0.0:12201"
    use_incoming_timestamp: true
    labels:
      job: gelf
  relabel_configs:
      - action: replace
        source_labels:
          - __gelf_message_host
        target_label: host
      - action: replace
        source_labels:
          - __gelf_message_level
        target_label: level
      - action: replace
        source_labels:
          - __gelf_message_facility
        target_label: facility

Cloudflare

Promtail supports pulling HTTP log messages from Cloudflare using the Logpull API. The Cloudflare targets can be configured with a cloudflare block:

scrape_configs:
- job_name: cloudflare
  cloudflare:
    api_token: REDACTED
    zone_id: REDACTED
    fields_type: all
    labels:
      job: cloudflare-foo.com

Only api_token and zone_id are required. Refer to the Cloudfare configuration section for details.

Heroku Drain

Promtail supports receiving logs from a Heroku application by using a Heroku HTTPS Drain. Configuration is specified in aheroku_drain block within the Promtail scrape_config configuration.

- job_name: heroku_drain
    heroku_drain:
      server:
        http_listen_address: 0.0.0.0
        http_listen_port: 8080
      labels:
        job: heroku_drain_docs
      use_incoming_timestamp: true
    relabel_configs:
      - source_labels: ['__heroku_drain_host']
        target_label: 'host'
      - source_labels: ['__heroku_drain_app']
        target_label: 'source'
      - source_labels: ['__heroku_drain_proc']
        target_label: 'proc'
      - source_labels: ['__heroku_drain_log_id']
        target_label: 'log_id'

Within the scrape_configs configuration for a Heroku Drain target, the job_name must be a Prometheus-compatible metric name.

The server section configures the HTTP server created for receiving logs. labels defines a static set of label values added to each received log entry. use_incoming_timestamp can be used to pass the timestamp received from Heroku.

Before using a heroku_drain target, Heroku should be configured with the URL where the Promtail instance will be listening. Follow the steps in Heroku HTTPS Drain docs for using the Heroku CLI with a command like the following:

heroku drains:add [http|https]://HOSTNAME:8080/heroku/api/v1/drain -a HEROKU_APP_NAME

Getting the Heroku application name

Note that the __heroku_drain_app label will contain the source of the log line, either app or heroku and not the name of the heroku application.

The easiest way to provide the actual application name is to include a query parameter when creating the heroku drain and then relabel that parameter in your scraping config, for example:

heroku drains:add [http|https]://HOSTNAME:8080/heroku/api/v1/drain?app_name=HEROKU_APP_NAME -a HEROKU_APP_NAME

And then in a relabel_config:

    relabel_configs:
      - source_labels: ['__heroku_drain_param_app_name']
        target_label: 'app'

It also supports relabeling and pipeline stages just like other targets.

When Promtail receives Heroku Drain logs, various internal labels are made available for relabeling:

__heroku_drain_host
__heroku_drain_app
__heroku_drain_proc
__heroku_drain_log_id In the example above, the project_id label from a GCP resource was transformed into a label called project through relabel_configs.

Relabeling

Each scrape_configs entry can contain a relabel_configs stanza. relabel_configs is a list of operations to transform the labels from discovery into another form.

A single entry in relabel_configs can also reject targets by doing an action: drop if a label value matches a specified regex. When a target is dropped, the owning scrape_config will not process logs from that particular source. Other scrape_configs without the drop action reading from the same target may still use and forward logs from it to Loki.

A common use case of relabel_configs is to transform an internal label such as __meta_kubernetes_* into an intermediate internal label such as __service__. The intermediate internal label may then be dropped based on value or transformed to a final external label, such as __job__.

Examples

Drop the target if a label (__service__ in the example) is empty:

  - action: drop
    regex: ''
    source_labels:
    - __service__

Drop the target if any of the source_labels contain a value:

  - action: drop
    regex: .+
    separator: ''
    source_labels:
    - __meta_kubernetes_pod_label_name
    - __meta_kubernetes_pod_label_app

Persist an internal label by renaming it so it will be sent to Loki:

  - action: replace
    source_labels:
    - __meta_kubernetes_namespace
    target_label: namespace

Persist all Kubernetes pod labels by mapping them, like by mapping __meta_kube__meta_kubernetes_pod_label_foo to foo.

  - action: labelmap
    regex: __meta_kubernetes_pod_label_(.+)

Additional reading:

Julien Pivotto’s slides from PromConf Munich, 2017

HTTP client options

Promtail uses the Prometheus HTTP client implementation for all calls to Loki. Therefore it can be configured using the clients stanza, where one or more connections to Loki can be established: