Sometimes when many rows can be returned a scalar index query might take some time. For example, a large range query against a btree index that hits most or all of the data. If there is a limit on the scan then we could potentially push this into the index itself so that we know we can short-circuit early.
Sometimes when many rows can be returned a scalar index query might take some time. For example, a large range query against a btree index that hits most or all of the data. If there is a limit on the scan then we could potentially push this into the index itself so that we know we can short-circuit early.