Given the number of data we have already ingested, we hit performance issues with package-releases job that checks new releases on registered indexes. The new implementation uses concurrent requests to the registered Python package indexes using asyncio to get information about releases faster.
https://github.com/thoth-station/pack...