log_to_metric in all_metrics mode dropping integer-valued metrics

[movinfinex]

A note for the community

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Problem

When using all_metrics with log_to_metric, metric values (as well as histogram upper limits, etc.) are extracted using as_float(): https://github.com/vectordotdev/vector/blob/bb6a91404cdc3f662b037cc4d5f52237b28d215d/src/transforms/log_to_metric.rs#L513-L522

However, if the log data serialiser writes the value as an integer, the code fails and the event is dropped with this error:

ERROR transform{component_kind="transform" component_id=metrics component_type=log_to_metric}: vector::internal_events::log_to_metric: Failed to parse field as float. field=counter.value error_code="failed_parsing_FloatError" error_type="parser_failed" stage="processing" internal_log_rate_limit=true

If the same value is serialised with a trailing .0, it works.

Configuration

sources:
  gen:
    type: "demo_logs"
    format: "shuffle"
    decoding:
      codec: "json"
    lines:
      - '{"counter":{"value":42},"kind":"incremental","name":"foo","tags":{}}' # "Failed to parse field as float."
      - '{"counter":{"value":42.0},"kind":"incremental","name":"foo","tags":{}}' # OK
transforms:
  metrics:
    type: "log_to_metric"
    inputs:
      - "gen"
    metrics: []
    all_metrics: true
sinks:
  console:
    type: "console"
    inputs:
      - "metrics"
    encoding:
      codec: "json"

Version

vector 0.47.0 (aarch64-apple-darwin 3d5af22 2025-05-20 13:53:41.638057046)

Debug Output

Example Data

No response

Additional Context

No response

References

No response