Sentinel-2 split tiles - data missing.

[mikkel-skyfi]

Sentinel-2 acquires long strips of data. The end of one strip, and the start of the next, has plenty of overlap, but does result in there being two different datasets for the same tile on some days.
To exemplify this, on scihub.copernicus.eu, this search:
( footprint:"Intersects(POLYGON((11.154589183358947 56.50664187308661,11.279520582212568 56.50664187308661,11.279520582212568 56.97644785826725,11.154589183358947 56.97644785826725,11.154589183358947 56.50664187308661)))" ) AND ( beginPosition:[2022-08-17T00:00:00.000Z TO 2022-08-17T23:59:59.999Z] AND endPosition:[2022-08-17T00:00:00.000Z TO 2022-08-17T23:59:59.999Z] ) AND ( (platformname:Sentinel-2 AND producttype:S2MSI2A))
Finds four images:

While a similar search against the Planetary Computer STAC interface nets me two images:

On average, it appears that this results in about 300 to 400 missing tiles / images per day.

[TomAugspurger]

Thanks for the report. I'll try to look into this this week. I know in the past we had reports that we were losing items thanks to what sounds like a similar issue (stactools-packages/sentinel2#5), which has since been fixed.