[ https://issues.apache.org/jira/browse/IGNITE-12715?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17042902#comment-17042902 ]
Roman Kondakov commented on IGNITE-12715: ----------------------------------------- [~Pavlukhin], yes, this is the most natural way to add indexes into the Calcite planner. > Calcite integration. Secondary indexes support. > ----------------------------------------------- > > Key: IGNITE-12715 > URL: https://issues.apache.org/jira/browse/IGNITE-12715 > Project: Ignite > Issue Type: New Feature > Components: sql > Reporter: Roman Kondakov > Assignee: Roman Kondakov > Priority: Major > > Secondary indexes should be supported by Calcite-based engine as well as they > supported by the legacy H2 engine. At first we can use the old engine for > indexes maintenance (building, updating, etc). In this case Calcite engine > will only use indexes metadata for query planning and index scans for query > execution. On the next iteration we need to eliminate the old engine usage. > Approximate plan for indexes support implementation: > # Add indexes to schema and set up all schema listeners. > # Add Collation to planner's output trait set and check if generated plan is > properly chosen in accordance to index sort direction. > # Implement index scans with filtering > # Add Sort node to exec. > -- This message was sent by Atlassian Jira (v8.3.4#803005)