[ https://issues.apache.org/jira/browse/IGNITE-14945?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17459185#comment-17459185 ]
Maksim Timonin commented on IGNITE-14945: ----------------------------------------- A lot of thanks to [~ibessonov] for suggestions related to IO part. Merged to master. > IndexQuery should use inline IO for internal filtering. > ------------------------------------------------------- > > Key: IGNITE-14945 > URL: https://issues.apache.org/jira/browse/IGNITE-14945 > Project: Ignite > Issue Type: New Feature > Reporter: Maksim Timonin > Assignee: Maksim Timonin > Priority: Major > Labels: IEP-71 > Time Spent: 20m > Remaining Estimate: 0h > > For comparison of index keys it's required: > # to init cache data row > # access fields with BinaryObject API > So, it's possible to use inline IO for filtering. It can help: > # speed up comparison (Inline IO access is faster than BinaryObject API). > # to avoid init cache data rows for filtered items (in case there are more > filtered items). -- This message was sent by Atlassian Jira (v8.20.1#820001)