MySQL Data Redistribution

Motivation

While B-Tree is a ubiquitous index structure used in managing data, it is well known for its low space utilization in nodes. Such space under-utilization is detrimental to flash storage in terms of cost and performance. In particular, the logical space waste in B-tree will amplify physical writes inside flash storage, worsening the transaction throughput.

Contribution

Our evaluation results from running OLTP benchmarks using the optimized MySQL/InnoDB prototype clearly show that those optimizations improve transaction throughput (i.e., more than 50%) with less space and cost (i.e., less than 40%) in flash storage.

Experiment Result

TPCC-Result
Index Space Utilization

Prerequisites & Installation Guide

Install prerequisites of mysql-5.6.26. Follow the instructions in the site.
Clone this repository.

$ git clone https://github.com/FlashSQL/MySQL-Data-Redistribution.git

Run mysqld server to run MySQL.

Modified files compared to Vanilla MySQL

storage/innobase/btr/btr0btr.cc
storage/innobase/fil/fil0fil.cc
storage/innobase/srv/srv0srv.cc
storage/innobase/include/btr0btr.h
storage/innobase/include/srv0srv.h
storage/innobase/include/fil0fil.h

References

https://dev.mysql.com/

About

Data Redistribution on MySQL

Other

Languages

Language:C++ 56.2%Language:C 32.0%Language:Java 3.6%Language:Perl 2.6%Language:Roff 1.1%Language:Yacc 0.7%Language:NASL 0.7%Language:CMake 0.7%Language:PHP 0.7%Language:Shell 0.6%Language:Objective-C 0.5%Language:Pascal 0.2%Language:Python 0.1%Language:GAP 0.1%Language:Makefile 0.1%Language:Assembly 0.0%Language:SWIG 0.0%Language:Lex 0.0%Language:HTML 0.0%Language:M4 0.0%Language:Scilab 0.0%Language:JavaScript 0.0%Language:Batchfile 0.0%Language:POV-Ray SDL 0.0%Language:Awk 0.0%Language:SourcePawn 0.0%Language:VBA 0.0%