geoparquet: Reading GeoParquet files fails by default on Azure due to failed suffix request

[kylebarron]

Should make an upstream issue for this and perhaps an upstream fix that will fall back to a head + range request in case the suffix request fails.

It's https://docs.rs/parquet/latest/parquet/arrow/arrow_reader/struct.ArrowReaderMetadata.html#method.load_async that fails on Azure if the file size wasn't provided.

In #1176 added an alternate API to pass in the file size alongside the file name