Global options reference

Controls how acknowledgements are handled for all sinks by default.

See End-to-end Acknowledgements for more information on how Vector handles event acknowledgement.

The directory used for persisting Vector state data.

This is the directory where Vector will store any state data, such as disk buffers, file checkpoints, and more.

Vector must have write permissions to this directory.

Configuration options for an enrichment table to be used in a remap transform. Currently supported are:

• CSV files
• MaxMind databases
• In-memory storage

For the lookup in the enrichment tables to be as performant as possible, the data is indexed according to the fields that are used in the search. Note that indices can only be created for fields for which an exact match is used in the condition. For range searches, an index isn’t used and the enrichment table drops back to a sequential scan of the data. A sequential scan shouldn’t impact performance significantly provided that there are only a few possible rows returned by the exact matches in the condition. We don’t recommend using a condition that uses only date range searches.

If set, Vector will configure the internal metrics system to automatically remove all metrics that have not been updated in the given time.

If set to a negative value expiration is disabled.

The amount of time, in seconds, that internal metrics will persist after having not been updated before they expire and are removed.

Set this to a value larger than your internal_metrics scrape interval (default 5 minutes) so metrics live long enough to be emitted and captured.

Default log schema for all events.

This is used if a component does not have its own specific log schema. All events use a log schema, whether or not the default is used, to assign event fields on incoming events.

Proxy configuration.

Configure to proxy traffic through an HTTP(S) proxy when making external requests.

Similar to common proxy configuration convention, you can set different proxies to use based on the type of traffic being proxied. You can also set specific hosts that should not be proxied.

Configuration options to retrieve secrets from external backend in order to avoid storing secrets in plaintext in Vector config. Multiple backends can be configured. Use SECRET[<backend_name>.<secret_key>] to tell Vector to retrieve the secret. This placeholder is replaced by the secret retrieved from the relevant backend.

When type is exec, the provided command will be run and provided a list of secrets to fetch, determined from the configuration file, on stdin as JSON in the format:

{"version": "1.0", "secrets": ["secret1", "secret2"]}

The executable is expected to respond with the values of these secrets on stdout, also as JSON, in the format:

{
    "secret1": {"value": "secret_value", "error": null},
    "secret2": {"value": null, "error": "could not fetch the secret"}
}

If an error is returned for any secrets, or if the command exits with a non-zero status code, Vector will log the errors and exit.

Otherwise, the secret must be a JSON text string with key/value pairs. For example:

{
    "username": "test",
    "password": "example-password"
}

If an error occurred while reading the file or retrieving the secrets, Vector logs the error and exits.

Secrets are loaded when Vector starts or if Vector receives a SIGHUP signal triggering its configuration reload process.

Telemetry options.

Determines whether source and service tags should be emitted with the component_sent_* and component_received_* events.

The name of the time zone to apply to timestamp conversions that do not contain an explicit time zone.

The time zone name may be any name in the TZ database or local to indicate system local time.

Note that in Vector/VRL all timestamps are represented in UTC.

Set wildcard matching mode for inputs

Setting this to “relaxed” allows configurations with wildcards that do not match any inputs to be accepted without causing an error.