Vector v0.47.0 release notes

Add wildcard_matching global config option to set wildcard matching mode for inputs. Relaxed mode allows configurations with wildcards that do not match any inputs to be accepted without causing an error.

Example config:

wildcard_matching: relaxed

sources:
  stdin:
    type: stdin

# note - no transforms

sinks:
  stdout:
    type: console
    encoding:
      codec: json
    inputs:
      - "runtime-added-transform-*"

Add an option in the datadog_logs sink to allow Vector to mutate the record to conform to the protocol used by the Datadog Agent itself. To enable, use the conforms_as_agent option or have the appropriate agent header (DD-PROTOCOL: agent-json) within the additional HTTP Headers list.

Any top-level fields that use Datadog-reserved keywords are moved into a new object named message. If message doesn’t exist, it is created first. For example:

{
  "key1": "value1",
  "key2": { "key2-1" : "value2" },
  "message" : "Hello world",
  ... rest of reserved fields
}

will be modified to:

{
  "message" : {
    "message" : "Hello world",
    "key1": "value1",
    "key2": { "key2-1" : "value2" }
  },
  ... rest of reserved fields
}

Add a new extra_headers option to greptimedb_logs sink configuration to set additional headers for outgoing requests.

Change greptimedb_logs sink default content type to application/x-ndjson to match the default content type of greptimedb sink. If you use the greptimedb version v0.12 or earlier, you need to set the content type to application/json in the sink configuration.

Example:

sinks:
  greptime_logs:
    type: greptimedb_logs
    inputs: ["my_source_id"]
    endpoint: "http://localhost:4000"
    table: "demo_logs"
    dbname: "public"
    extra_headers:
      x-source: vector

[sinks.greptime_logs]
type = "greptimedb_logs"
inputs = ["my_source_id"]
endpoint = "http://localhost:4000"
table = "demo_logs"
dbname = "public"

[sinks.greptime_logs.extra_headers]
x-source = "vector"