[CASSANDRA-15215] VIntCoding should read and write more efficiently

• Type: Bug
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Fix Version/s: 4.1-alpha1, 4.1
• Component/s: Local/Compaction, Local/SSTable
• Bug Category: Degradation - Performance Bug/Regression
• Severity: Normal
• Complexity: Normal
• Discovered By: Code Inspection
• Platform: All
• Since Version: 4.1

Source Control Link

35dbcc2c2dbe1c826fd6ecd6e8357f0f5a9bab02, 5cd012736e4680bbb25928ee7fbcea4859878fff

Test and Documentation Plan

Added unit tests for new methods and benchmarks to show performance improvements.

• Assignee: Alex Sorokoumov
• Reporter: Benedict Elliott Smith

Authors

Alex Sorokoumov

Reviewers

Benedict Elliott Smith, Branimir Lambov

• Created: 16/Jul/19 09:36
• Updated: 11/Nov/22 19:00
• Resolved: 02/Feb/22 12:20

Description

Most vints occupy significantly fewer than 8 bytes, and most buffers have >= 8 bytes spare, in which case we can construct the relevant bytes in a register and memcpy them to the correct position. Since we read and write a lot of vints, this waste is probably measurable, particularly during compaction and flush, and can probably be considered a performance bug.