Skip to main content
Vector is configured using a single configuration file that defines sources, transforms, sinks, and global options. This flexible configuration system allows you to build powerful observability data pipelines.

​
Configuration Formats

Vector supports multiple configuration formats:
  • YAML (recommended) - Human-readable and widely used
  • TOML - Simple and expressive
  • JSON - Machine-friendly format

​
Basic Configuration Example

Here’s a simple configuration that reads from stdin and outputs to console: 
sources:
  in:
    type: stdin

sinks:
  out:
    inputs: [in]
    type: console
    encoding:
      codec: text

​
Configuration Structure

A Vector configuration consists of four main sections:

​
1. Global Options

Global settings that apply to the entire Vector instance: 
data_dir: "/var/lib/vector"

log_schema:
  host_key: "host"
  message_key: "message"
  timestamp_key: "timestamp"

​
2. Sources

Data ingestion components that collect events from various sources: 
sources:
  apache_logs:
    type: file
    include: ["/var/log/apache2/*.log"]
    ignore_older: 86400

​
3. Transforms

Data processing components that modify, filter, or enrich events: 
transforms:
  parse_logs:
    type: remap
    inputs: [apache_logs]
    source: |
      . = parse_apache_log!(.message)

​
4. Sinks

Data output components that send events to destinations: 
sinks:
  elasticsearch:
    type: elasticsearch
    inputs: [parse_logs]
    endpoint: "http://localhost:9200"
    bulk:
      index: "vector-%Y-%m-%d"

​
Complete Example

Here’s a complete configuration that demonstrates all components: 
# Global options
data_dir: "/var/lib/vector"

log_schema:
  host_key: "host"
  message_key: "message"
  timestamp_key: "timestamp"

# Data ingestion
sources:
  apache_logs:
    type: file
    include: ["/var/log/apache2/*.log"]
    ignore_older: 86400

# Data processing
transforms:
  parse_logs:
    type: remap
    inputs: [apache_logs]
    source: |
      . = parse_apache_log!(.message)
  
  sample_logs:
    type: sample
    inputs: [parse_logs]
    rate: 2

# Data output
sinks:
  elasticsearch:
    type: elasticsearch
    inputs: [sample_logs]
    endpoint: "http://localhost:9200"
    bulk:
      index: "vector-%Y-%m-%d"
  
  s3_archive:
    type: aws_s3
    inputs: [parse_logs]
    region: us-east-1
    bucket: log-archives
    compression: gzip
    encoding:
      codec: json

​
Configuration Loading

Vector loads configuration files in order and merges them together. You can specify multiple configuration files: 
vector --config /etc/vector/vector.yaml --config /etc/vector/sources.yaml

​
Environment Variables

Vector supports environment variable interpolation in configuration files: 
sources:
  demo_logs:
    type: demo_logs
    format: "${LOG_FORMAT:-syslog}"
    interval: ${LOG_INTERVAL:-1}

​
Configuration Validation

Validate your configuration before starting Vector: 
vector validate --config /etc/vector/vector.yaml

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us