Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Done
-
None
-
None
Description
This change is regarding introducing secondary primary index. The secondary primary index is a secondary B-Tree that is built on the primary keys of the dataset but only stores the
{PK}instead of the regular case
{SK,PK}. The purpose of this secondary is to optimize queries like SELECT count. Instead of scanning the primary index, the secondary primary index would be used to answer such queries.