[SPARK-26780] Improve shuffle read using ReadAheadInputStream

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Won't Fix
• Affects Version/s: 3.0.0
• Component/s: Shuffle, Spark Core
• Assignee: Unassigned
• Reporter: liuxian
• Created: 30/Jan/19 07:36
• Updated: 17/May/20 18:30
• Resolved: 02/Jan/20 13:32

Description

Using ReadAheadInputStream  to improve  shuffle read  performance.
 ReadAheadInputStream can save cpu utilization and almost no performance regression

Issue Links

• links to: GitHub Pull Request #23695