Once T350863: Create maintenance script to import all existing images to mediamoderation_scan table and T350323: Write an empty row to scan table on file upload is complete, the existing images on WMF wikis will need to be imported to the mediamoderation_scan table. This task tracks the progress for this.
This will need to wait until either:
We ran this on testwiki today successfully.
@Dreamy_Jazz @kostajh Should we try to run this on all production wikis asap, assuming we intend to do that manually? (I know we're waiting a bit before we run the scanning script on the rest of production.)
Yes. The import on commons will likely take a while, so we should do this asap at least for commons so we could start a scan on commons over the holidays.
ImportExistingFilesToScanTable was run on group0 wikis today (took around 10 minutes). Once group1 are on 1.42-wmf.10 we'll run it on those.
Import script was run on all wikis. The number of rows in mediamoderation_scan is sometimes lower than the sum of rows in image + oldimage + filearchive, as expected: some file types can't be scanned; some SHA-1s are duplicated.
Leaving open in case there's any follow-up work to do.