You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
DuckLake's metadata-only COUNT(*) optimization, introduced in, recomputes the delete-adjusted row count whenever partition statistics are requested.
This executes one query for data-file rows adjusted by deletes and another for inlined rows, even when the same table is queried again at the same snapshot. The optimization already excludes time-travel queries and tables with transaction-local changes.
Could we cache the net row count by table and snapshot, then recompute it only when the snapshot changes?
This would remove repeated metadata queries without changing COUNT(*).
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
DuckLake's metadata-only
COUNT(*)optimization, introduced in, recomputes the delete-adjusted row count whenever partition statistics are requested.This executes one query for data-file rows adjusted by deletes and another for inlined rows, even when the same table is queried again at the same snapshot. The optimization already excludes time-travel queries and tables with transaction-local changes.
Could we cache the net row count by table and snapshot, then recompute it only when the snapshot changes?
This would remove repeated metadata queries without changing
COUNT(*).All reactions