README.md

Awesome Database Learning

A list of learning materials to understand databases internals, including but not limited to:

• papers
• blogs
• courses
• talks

Please submit a pull request if there is any material that you think should be included in this collection.

Table of Contents

• Recommended Courses and Books
  • Courses
  • Books
• SQL & Relation Algebra
• Query Optimizer
  • Planner Models
  • Subquery Optimization
  • Join Order Optimization
  • Functional Dependency & Physical Properties
  • Cost Model
  • Statistics
• Query Execution
  • Execution Framework
  • Vectorization vs Compilization
  • Join
  • Hash Table
• DDL
• Transaction
  • Isolation Levels
  • Concurrency Control
• Network
• Storage
  • Buffer Management
  • Disk IO
  • B-Tree
  • LSM-Tree
• Serializing & RPC
• Data Partitioning
• Replication & Consistency
• Consensus
• Scale & Balance
• Benchmark & Testing

Recommended Courses and Books

Courses

• CMU Database Systems (15-445/645), thanks to Andy Pavlo
• CMU Advanced Database Systems (15-721), thanks to Andy Pavlo
• UC Berkeley Introduction to Database Systems

Books

• Stanford Database Systems: The Complete Book
• Designing Data-Intensive Applications, 中文翻译
• Database Internals

SQL & Relation Algebra

Courses:

• CMU Database Systems (15-445/645), thanks to Andy Pavlo

  • Course Introduction and the Relational Model
  • Advanced SQL

• UC Berkeley Introduction to Database Systems

  • Introduction + SQL I
  • SQL II
  • Relational Algebra

Query Optimizer

Blogs:

• 数据库内核杂谈, thanks to 顾仲贤
  • 数据库内核杂谈（七）：数据库优化器（上）
  • 数据库内核杂谈（八）：数据库优化器（下）
• SQL优化器原理 - 查询优化器综述, thanks to 勿烦

Planner Models

Blogs:

• 数据库内核杂谈, thanks to 顾仲贤
  • 数据库内核杂谈（九）：开源优化器 ORCA
• SQL 查询优化原理与 Volcano Optimizer 介绍, thanks to 张茄子
• Cascades Optimizer, thanks to hellocode

Papers:

• 1979, Access Path Selection in a Relational Database Management System, SIGMOD
• 1979, Query Processing in Main Memory Database Management Systems, VLDB
• 1987, Query Optimization by Simulated Annealing, SIGMOD
• 1988, Grammar-like Functional Rules for Representing Query Optimization Alternatives, SIGMOD
• 1993, The Volcano Optimizer Generator- Extensibility and Efficient Search, ICDE
• 1995, The Cascades Framework for Query Optimization, IEEE Data engineering Bulltin
• 1998, An Overview of Query Optimization in Relational Systems, PODS
• 2001, LEO – DB2’s LEarning Optimizer, VLDB
• 2004, Robust Query Processing through Progressive Optimization, SIGMOD
• 2014, Orca: A Modular Query Optimizer Architecture for Big Data, SIGMOD
• 2016, Parallelizing Query Optimization on Shared-Nothing Architectures, VLDB
• 2016, The MemSQL Query Optimizer: A modern optimizer for real-time analytics in a distributed database, VLDB

Subquery Optimization

Blogs:

• SQL 子查询的优化, thanks to Eric Fu
• Calcite 子查询处理 - I (RemoveSubQuery), thanks to 一只无情的小猫咪
• Calcite 子查询处理 - II (Decorrelate), thanks to 一只无情的小猫咪

Papers:

• 2001, Orthogonal Optimization of Subqueries and Aggregation, SIGMOD
• 2009, Enhanced subquery optimizations in Oracle, VLDB
• 2015, Unnesting Arbitrary Queries, BTW

Join Order Optimization

Papers:

• 2006, Analysis of Two Existing and One New Dynamic Programming Algorithm for the Generation of Optimal Bushy Join Trees without Cross Products, VLDB
• 2015, How Good Are Query Optimizers, Really?, VLDB
• 2018, Adaptive Optimization of Very Large Join Queries, SIGMOD

Functional Dependency & Physical Properties

Thesis:

• 2000, Exploiting Functional Dependence in Query Optimization

Papers:

• 1996, Fundamental Techniques for Order Optimization, SIGMOD
• 2004, An Efficient Framework for Order Optimization, ICDE
• 2010, Incorporating Partitioning and Parallel Plans into the SCOPE Optimizer, ICDE

Cost Model

Papers:

• 1996, Modelling Costs for a MM-DBMS, in Real-Time Databases
• 2014, Approximation Schemes for Many-Objective Query Optimization, SIGMOD
• 2015, Multi-Objective Parametric Query Optimization, VLDB

Statistics

Papers:

• 2003, The History of Histograms, VLDB
• 2005, An Improved Data Stream Summary: The Count-Min Sketch and its Applications, Journal of Algorithms
• 2007, New Estimation Algorithms for Streaming Data: Count-min Can Do More
• 2017, Adaptive Statistics in Oracle 12c, VLDB

Books:

• Synopses for Massive Data: Samples, Histograms, Wavelets, Sketches

Query Execution

Execution Framework

Papers:

• 1994, Volcano-An Extensible and Parallel Query Evaluation System, IEEE Transactions on Knowledge and Data EngineeringFebruary
• 2014, Morsel-Driven Parallelism: A NUMA-Aware Query Evaluation Framework for the Many-Core Age, SIGMOD

Vectorization vs Compilization

Blogs:

• Overhead of a Generalized Query Execution Engine, from The Pivotal Engineering Journal, thanks to the Pivotal Engineering team

Papers:

• 2005, MonetDB/X100: Hyper-Pipelining Query Execution, CIDR
• 2011, Efficiently Compiling Efficient Query Plans for Modern Hardware, VLDB
• 2017, Relaxed Operator Fusion for In-Memory Databases: Making Compilation, Vectorization, and Prefetching Work Together At Last, VLDB
• 2018, Everything You Always Wanted to Know About Compiled and Vectorized Queries But Were Afraid to Ask, VLDB
• 2018, Adaptive Execution of Compiled Queries, ICDE

Join

Papers:

• 2013, Multi-Core, Main-Memory Joins: Sort vs. Hash Revisited, VLDB
• 2017, Looking Ahead Makes Query Plans Robust, VLDB

Hash Table

Blogs:

• Fibonacci Hashing: The Optimization that the World Forgot (or: a Better Alternative to Integer Modulo), thanks to Malte Skarupke
• All hash table sizes you will ever need, thanks to Database Architects - Thomas Neumann

DDL

• 2013, Online, Asynchronous Schema Change in F1, VLDB

Transaction

Isolation Levels

Blogs:

• 一致性模型, thanks to siddontang

Papers:

• 1995, A Critique of ANSI SQL Isolation Levels, SIGMOD

Concurrency Control

Courses:

• CMU Database Systems (15-445/645), thanks to Andy Pavlo

  • Concurrency Control Theory
  • Two-Phase Locking Concurrency Control
  • Timestamp Ordering Concurrency Control
  • Multi-Version Concurrency Control

• CMU Advanced Database Systems (15-721), thanks to Andy Pavlo

  • Multi-Version Concurrency Control (Design Decisions)
  • Multi-Version Concurrency Control (Protocols)
  • Multi-Version Concurrency Control (Garbage Collection)

Papers:

• 2012, Serializable Snapshot Isolation in PostgreSQL, VLDB
• 2015, Fast Serializable Multi-Version Concurrency Control for Main-Memory Database Systems, SIGMOD
• 2017, An Empirical Evaluation of In-Memory Multi-Version Concurrency Control, VLDB
• 2019, Scalable Garbage Collection for In-Memory MVCC Systems, VLDB

Network

Courses:

• CMU Advanced Database Systems (15-721), thanks to Andy Pavlo
  • Networking Protocols

Papers:

• 2016, The End of Slow Networks: It's Time for a Redesign, VLDB
• 2016, Accelerating Relational Databases by Leveraging Remote Memory and RDMA, SIGMOD
• 2017, Don't Hold My Data Hostage: A Case for Client Protocol Redesign, VLDB

Storage

Buffer Management

Courses:

• CMU Database Systems (15-445/645), thanks to Andy Pavlo
  • Buffer Pools

Papers:

• 1987, The 5 Minute Rule for Trading Memory for Disc Accesses and the 5 Byte Rule for Trading Memory for CPU Time, SIGMOD
• 2008, The Five Minute Rule 20 Years Later and How Flash Memory Changes the Rules, ACM Queue
• 2018, Managing Non-Volatile Memory in Database Systems, SIGMOD
• 2018, LeanStore: In-Memory Data Management Beyond Main Memory, ICDE
• 2020, Umbra: A Disk-Based System with In-Memory Performance, CIDR

Disk IO

Blogs:

• On Disk IO, Part 1: Flavors of IO, thanks to Alex
• On Disk IO, Part 2: More Flavours of IO, thanks to Alex
• Read, write & space amplification - pick 2, thanks to Mark Callaghan

Papers:

• 2016, Design Tradeoffs of Data Access Methods, SIGMOD
• 2016, Designing Access Methods: The RUM Conjecture, EDBT

B-Tree

Courses:

• CMU Advanced Database Systems (15-721), thanks to Andy Pavlo
  • OLTP Indexes (B+Tree Data Structures)

LSM-Tree

Serializing & RPC

Data Partitioning

Replication & Consistency

Consensus

• University of Cambridge Distributed consensus revised, a great paper about Consenssus especially Paxos and Paxos-Related algorithms, by Heidi Howard

Scale & Balance

Blogs:

• Building a Large-scale Distributed Storage System Based on Raft, by Ed Huang

Benchmark & Testing