Get series data in various formats based on query parameter.

UUID of the series

Output format: senml, csv, jsonl or arrow (default: senml)

Start datetime in ISO 8601 format (e.g., '2024-01-15T10:30:00Z')

End datetime in ISO 8601 format (e.g., '2024-01-15T11:00:00Z')

Maximum number of samples (at most 100,000). Returns the oldest N samples of the window, in time order, and silently drops the rest. With simplify, it bounds the rows read before simplification

Bucket width using Prometheus duration syntax (e.g., '1h', '5m', '1d')

Aggregation function: avg, min, max, sum, count, first, last

Explicitly enable simplify-based point reduction. Applied to the rows read for the window (after bucketing when step is set); more than 100,000 rows to simplify is rejected with HTTP 400 instead of being truncated, so use step/aggregation or a narrower start/end

Dimensionless simplify tolerance on normalized time/value coordinates

Use Douglas-Peucker only when simplifying

Answer with 'Content-Disposition: attachment' and a file name: the series name, its labels, the window and the step (e.g. 'temperature_room-kitchen_20261006T090000Z_20261006T100000Z_5m-avg.csv'), so that a browser or 'curl -OJ' saves a file