If DB2 used a single index rangescan from 20 to 30, it would miss this document and return an incomplete query result.
如果DB 2在20到30之间使用单个索引范围扫描,则会错过该文档,并返回不完全的查询结果。
2
If your intention is a real "between" predicate, then the execution plan on the right is much better, because a single rangescan with a start-top predicate delivers the matching row IDs only.