You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@ignite.apache.org by "Konstantin Orlov (Jira)" <ji...@apache.org> on 2022/01/13 13:59:00 UTC

[jira] [Created] (IGNITE-16296) Sql. Refactoring of integration with Data layer

Konstantin Orlov created IGNITE-16296:
-----------------------------------------

             Summary: Sql. Refactoring of integration with Data layer
                 Key: IGNITE-16296
                 URL: https://issues.apache.org/jira/browse/IGNITE-16296
             Project: Ignite
          Issue Type: Improvement
          Components: sql
            Reporter: Konstantin Orlov
            Assignee: Konstantin Orlov


Currently, the SQL layer works with the SqlRow abstract type (which is actually a generic type), while the data layer works with BinaryRow. The Tuple was chosen as the intermediary object because of its simplicity. But this brings one more transformation for each row. Besides, resolving a tuple from a BinaryRow requires interaction with the metadata store to keep the schema in sync.

All this renders a very slow execution of SQL operations. So I suggest reworking the integration to directly convert SqlRow to BinaryRow and vice versa.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)