In most instances, we’ve observed that only a subset of byte ranges are never accessed, rather than entire files being read (either in chunks or as a whole). Adapting to this, the current algorithm caches the entire file, anticipating that additional chunks will be requested in the future. This data movement happens in the background to prevent any latency issues in the request path, i.e., the content‑range request will not be delayed while the file is being moved.