Skip to main content
Version: Next

InfluxDB

The InfluxDB output component writes time-series data to an InfluxDB 2.x server using the Line Protocol. It supports batching, automatic field type detection, and configurable retry logic.

Configuration​

url​

InfluxDB server URL.

type: string

example: "http://localhost:8086"

optional: false

org​

Organization name in InfluxDB 2.x.

type: string

optional: false

bucket​

Bucket name to write data to.

type: string

optional: false

token​

Authentication token for InfluxDB 2.x API.

type: string

optional: false

measurement​

Measurement name for the time-series data.

type: string

optional: false

tags (optional)​

Tag mappings for indexed fields. Tags are indexed in InfluxDB and used for fast queries.

type: array of object

optional: true

tags[].name​

Tag name in InfluxDB.

type: string

optional: false (when tags is specified)

tags[].value​

Column name in the data to map to the tag.

type: string

optional: false (when tags is specified)

fields​

Field mappings for value fields. Fields contain the actual data values.

type: array of object

optional: false

fields[].name​

Field name in InfluxDB.

type: string

optional: false

fields[].value​

Column name in the data to map to the field.

type: string

optional: false

fields[].value_type​

Data type of the field.

Supported values: float, integer, boolean, string

type: string

optional: false

timestamp_field (optional)​

Column name to use as the timestamp. If not specified, InfluxDB will use the current time.

type: string

optional: true

batch_size (optional)​

Number of points to batch before writing to InfluxDB.

type: integer

default: 1000

optional: true

flush_interval (optional)​

Maximum time to wait before flushing batched data.

type: string

default: "5s"

optional: true

retry_count (optional)​

Number of retry attempts for failed writes.

type: integer

default: 3

optional: true

retry_delay (optional)​

Initial delay between retries in milliseconds.

type: integer

default: 1000

optional: true

Examples​

Basic InfluxDB Output​

- output:
    type: "influxdb"
    url: "http://localhost:8086"
    org: "my-org"
    bucket: "sensor-data"
    token: "my-token"
    measurement: "temperature"
    fields:
      - name: "value"
        value: "temperature"
        value_type: "float"
      - name: "sensor"
        value: "sensor_id"
        value_type: "string"

With Tags and Timestamp​

- output:
    type: "influxdb"
    url: "http://localhost:8086"
    org: "production"
    bucket: "metrics"
    token: "${INFLUXDB_TOKEN}"
    measurement: "system_metrics"
    tags:
      - name: "location"
        value: "datacenter"
      - name: "host"
        value: "hostname"
      - name: "region"
        value: "region"
    fields:
      - name: "cpu_usage"
        value: "cpu_percent"
        value_type: "float"
      - name: "memory_usage"
        value: "memory_mb"
        value_type: "integer"
      - name: "is_active"
        value: "active"
        value_type: "boolean"
      - name: "status"
        value: "status_message"
        value_type: "string"
    timestamp_field: "timestamp"

With Batching and Retry​

- output:
    type: "influxdb"
    url: "https://influxdb.example.com:8086"
    org: "enterprise"
    bucket: "telemetry"
    token: "${INFLUXDB_TOKEN}"
    measurement: "iot_readings"
    tags:
      - name: "device_type"
        value: "device_type"
      - name: "location"
        value: "location"
    fields:
      - name: "temperature"
        value: "temp"
        value_type: "float"
      - name: "humidity"
        value: "humidity"
        value_type: "float"
      - name: "battery_level"
        value: "battery"
        value_type: "integer"
      - name: "online"
        value: "is_online"
        value_type: "boolean"
    timestamp_field: "event_time"
    batch_size: 5000
    flush_interval: 10s
    retry_count: 5
    retry_delay: 2000

Sensor Data Example​

- output:
    type: "influxdb"
    url: "http://influxdb:8086"
    org: "manufacturing"
    bucket: "sensor-data"
    token: "${INFLUXDB_TOKEN}"
    measurement: "production_line"
    tags:
      - name: "line"
        value: "production_line"
      - name: "machine"
        value: "machine_id"
      - name: "shift"
        value: "shift_id"
    fields:
      - name: "temperature"
        value: "temp_celsius"
        value_type: "float"
      - name: "pressure"
        value: "pressure_psi"
        value_type: "float"
      - name: "vibration"
        value: "vibration_hz"
        value_type: "float"
      - name: "cycle_count"
        value: "cycles"
        value_type: "integer"
      - name: "maintenance_required"
        value: "needs_maintenance"
        value_type: "boolean"
      - name: "operator"
        value: "operator_name"
        value_type: "string"
    timestamp_field: "reading_time"
    batch_size: 1000
    flush_interval: 5s

Features​

  • Line Protocol: Native InfluxDB 2.x Line Protocol format
  • Automatic Escaping: Special characters in tags and fields are automatically escaped
  • Type Support: Float, Integer, Boolean, and String field types
  • Batching: Efficient batch writes with configurable batch size and flush interval
  • Retry Logic: Automatic retry with exponential backoff for failed writes
  • Timestamp Handling: Optional timestamp field for precise time-series data
  • Tags and Fields: Proper separation between indexed tags and value fields

Best Practices​

  1. Use Tags for Filtering: Tags are indexed in InfluxDB. Use them for fields you frequently filter or group by.
  2. Use Fields for Values: Store actual measurements and values in fields.
  3. Optimize Batch Size: Larger batch sizes (1000-10000) generally improve performance but increase memory usage.
  4. Set Appropriate Flush Interval: Balance between latency and throughput with flush_interval.
  5. Timestamp Precision: Ensure your timestamp field is in nanoseconds for InfluxDB precision.
  6. Cardinality Management: Be careful with high-cardinality tags (tags with many unique values).

Field Types​

Float​

Floating-point numbers. Use for measurements with decimal precision.

- name: "temperature"
  value: "temp"
  value_type: "float"

Integer​

Whole numbers. Use for counts and discrete values.

- name: "count"
  value: "event_count"
  value_type: "integer"

Boolean​

True/false values.

- name: "active"
  value: "is_active"
  value_type: "boolean"

String​

Text values.

- name: "status"
  value: "status_message"
  value_type: "string"

Error Handling​

The output component will:

  • Retry failed writes up to retry_count times
  • Use exponential backoff starting at retry_delay milliseconds
  • Log errors for debugging
  • Continue processing on non-fatal errors