[ARROW-5197] [Java] Improving Arrow Vector Reading performance

• Type: Improvement
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Component/s: Java
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/21672
• Assignee: Unassigned
• Reporter: Yurui Zhou
• Created: 23/Apr/19 04:16
• Updated: 11/Jan/23 07:38

Description

Currently the read interface of Java Arrow Vector is quite slow because the access operation has to go through validity bit check and boundary check before it can actually load the data. Such a safety check incurs unnecessary overhead in actually computation. 

The Arrow Vector and ArrowBuf should expose unsafe methods for advanced users to directly access underlying data without null check and boundary check.

Issue Links

• is related to: ARROW-5200 [Java] Provide light-weight arrow APIs
• relates to: ARROW-1833 [Java] Add accessor methods for data buffers that skip null checking
• links to: GitHub Pull Request #4199

Sub-Tasks

• [Java] Add hasNull flag to Vectors: Open
• [Java] Add unsafe access method to ArrowBuf: Closed