snazy commented on PR #433: URL: https://github.com/apache/polaris/pull/433#issuecomment-2497422885
I think, we should have a design about how to handle table/view/udf metadata and its snapshots/versions and schemas/specs/etc and probably a vision how to deal with manifest list/file information. Handling these metadata objects as "opaque blobs" is quite expensive (no database is made to handle such huge objects). It would be nice to be have the information in a more structure way, but also efficiently accessible. Some use cases I can think of are: * For queries: just return the latest snapshot / just return one specific snapshot * Automatic metadata cleanup (remove unreferenced schema/spec/etc) With this change, we exchange the (costly) round-trip to the object store with another (costly) round-trip to a database. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
