Skip to main content
Global options are top-level configuration settings that apply to your entire Vector instance. These settings control fundamental behaviors like data storage, logging, and acknowledgements.

​
Configuration

Global options are configured at the top level of your configuration file: 
# Global options
data_dir: "/var/lib/vector"

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

acknowledgements:
  enabled: true

# Component configuration
sources:
  # ...

​
Data Directory

The directory where Vector stores state data, including disk buffers, file checkpoints, and other persistent data.
​
data_dir
string
default:"/var/lib/vector"
Path to the directory for persisting Vector state data. Vector must have write permissions to this directory.
data_dir: "/var/lib/vector"

​
Usage

The data directory is used for:
  • Disk buffers - When sinks use disk buffering
  • File checkpoints - Track file reading positions
  • State data - Component-specific state persistence

​
Permissions

Vector requires:
  • Read and write access to the data directory
  • Ability to create subdirectories
  • Sufficient disk space for buffers and state

​
Log Schema

Defines the default field names for core event data across all components.
​
log_schema
object
Default log schema for all events. Components can override these settings.
​
log_schema.host_key
string
default:"host"
Field name for the host that generated the event.
log_schema:
  host_key: "hostname"
​
log_schema.message_key
string
default:"message"
Field name for the event message/content.
log_schema:
  message_key: "msg"
​
log_schema.timestamp_key
string
default:"timestamp"
Field name for the event timestamp.
log_schema:
  timestamp_key: "ts"
​
log_schema.source_type_key
string
default:"source_type"
Field name for the source type identifier.
log_schema:
  source_type_key: "source"
​
log_schema.metadata_key
string
default:"metadata"
Field name for event metadata.
log_schema:
  metadata_key: "meta"

​
Example

log_schema:
  host_key: "server"
  message_key: "log_message"
  timestamp_key: "event_time"
  source_type_key: "log_source"
  metadata_key: "event_metadata"

sources:
  demo:
    type: demo_logs
    format: json

​
Acknowledgements

Control how Vector handles end-to-end acknowledgements for event delivery.
​
acknowledgements
object
Global acknowledgement configuration. Can be overridden per-sink.
​
acknowledgements.enabled
boolean
default:"false"
Enable acknowledgements globally for all sinks.
acknowledgements:
  enabled: true
When acknowledgements are enabled:
  • Sources wait for sinks to confirm event delivery
  • Prevents data loss during failures
  • Adds latency to event processing
  • Requires sink support for acknowledgements
See the End-to-end Acknowledgements documentation for details.

​
Timezone

Set the default timezone for timestamp operations.
​
timezone
string
default:"local"
Default timezone for timestamp conversions. Use any TZ database name or local for system time.
timezone: "America/New_York"
Common timezone values:
  • UTC - Coordinated Universal Time
  • America/New_York - US Eastern Time
  • Europe/London - UK Time
  • Asia/Tokyo - Japan Time
  • local - System local time

​
Proxy Configuration

Configure HTTP/HTTPS proxy settings for components that make external requests.
​
proxy
object
Global proxy configuration.
​
proxy.http
string
HTTP proxy URL.
proxy:
  http: "http://proxy.example.com:3128"
​
proxy.https
string
HTTPS proxy URL.
proxy:
  https: "http://proxy.example.com:3128"
​
proxy.no_proxy
array
List of domains to exclude from proxying.
proxy:
  no_proxy:
    - localhost
    - 127.0.0.1
    - "*.internal.example.com"

​
Example

proxy:
  http: "http://proxy.example.com:3128"
  https: "http://proxy.example.com:3128"
  no_proxy:
    - localhost
    - 127.0.0.1
    - "*.local"

​
Metrics Expiration

Control how long internal metrics persist before expiring.
​
expire_metrics_secs
float
default:"300"
Seconds before internal metrics expire after not being updated. Set higher than your metrics scrape interval.
expire_metrics_secs: 600  # 10 minutes

​
Healthchecks

Configure healthcheck behavior for all sinks.
​
healthchecks
object
Global healthcheck configuration.
​
healthchecks.enabled
boolean
default:"true"
Enable healthchecks for all sinks.
healthchecks:
  enabled: true
  require_healthy: false
​
healthchecks.require_healthy
boolean
default:"false"
Exit during startup if any sink healthcheck fails.
healthchecks:
  require_healthy: true

​
Enrichment Tables

Define enrichment tables for data augmentation.
​
enrichment_tables
object
Named enrichment tables that can be used by transforms.

​
Example

enrichment_tables:
  geoip:
    type: geoip
    path: "/etc/vector/GeoLite2-City.mmdb"

transforms:
  enrich:
    type: remap
    inputs: [logs]
    source: |
      .geo = get_enrichment_table_record!("geoip", {"ip": .client_ip})

​
Complete Example

# Global Options
data_dir: "/var/lib/vector"

log_schema:
  host_key: "hostname"
  message_key: "log"
  timestamp_key: "ts"
  source_type_key: "source"

timezone: "America/New_York"

proxy:
  http: "http://proxy.internal:3128"
  https: "http://proxy.internal:3128"
  no_proxy:
    - localhost
    - "*.internal"

acknowledgements:
  enabled: true

healthchecks:
  enabled: true
  require_healthy: true

expire_metrics_secs: 600

# Enrichment Tables
enrichment_tables:
  user_data:
    type: file
    file:
      path: "/etc/vector/users.csv"
      encoding:
        type: csv
    schema:
      user_id: string
      name: string
      email: string

# Sources
sources:
  app_logs:
    type: file
    include: ["/var/log/app/*.log"]

# Transforms
transforms:
  parse:
    type: remap
    inputs: [app_logs]
    source: |
      . = parse_json!(.log)
      .user = get_enrichment_table_record!("user_data", {"user_id": .user_id})

# Sinks
sinks:
  elasticsearch:
    type: elasticsearch
    inputs: [parse]
    endpoint: "http://localhost:9200"
    bulk:
      index: "logs-%Y-%m-%d"

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us