docs: add dbms research documentation

src/dbtest/src/doc/Database_Research_Doc_en.md

@@ -0,0 +1,78 @@
+### DBMS Overview
+
+| DBMS Name            | DBMS Type        | ODBC Support | Transaction Support | 3TS-Coo Verified | Supported Isolation Levels                                   | Concurrency Control Algorithms and Transaction Details |
+| -------------------- | ---------------- | ------------ | ------------------- | ---------------- | ------------------------------------------------------------ | ------------------------------------------------------ |
+| MySQL                | Relational DBMS  | Yes          | Yes                 | Yes              | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC, Optimistic and Pessimistic Control               |
+| PostgreSQL           | Relational DBMS  | Yes          | Yes                 | Yes              | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC                                                   |
+| MariaDB              | Relational DBMS  | Yes          | Yes                 | Yes              | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC                                                   |
+| Microsoft SQL Server | Relational DBMS  | Yes          | Yes                 | Yes              | Read Uncommitted, Read Committed, Repeatable Read, Serializable, Snapshot | 2PL, MVCC, Optimistic and Pessimistic Control          |
+| Oracle Database      | Relational DBMS  | Yes          | Yes                 | Yes              | Read Committed, Serializable                                 | 2PL (No-wait, Wait-for-Grant), MVCC                    |
+| MyRocks              | Relational DBMS  | Yes          | Yes                 | Yes              | Read Committed, Repeatable Read                              | MVCC                                                   |
+| MongoDB              | NoSQL DBMS       | Yes          | Supported since 3.0 | Yes              | Read Uncommitted, Snapshot                                   | Optimistic Concurrency Control (OCC)                   |
+| Redis                | NoSQL DBMS       | No           | No                  | No               | Not applicable                                               | Optimistic Lock, MULTI/EXEC                            |
+| Cassandra            | NoSQL DBMS       | Yes          | Yes                 | Yes              | Not applicable (controlled by consistency levels)            | LWT, Timestamp Ordering Protocol (T/O)                 |
+| Neo4j                | Graph Database   | Yes          | Yes                 | No               | Read Committed                                               | MVCC                                                   |
+| InfluxDB             | Time Series DBMS | Yes          | Yes                 | No               | Not applicable                                               | Optimistic Concurrency Control (OCC)                   |
+| TimescaleDB          | Time Series DBMS | Yes          | Yes                 | No               | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC                                                   |
+| CockroachDB          | Distributed DBMS | Yes          | Yes                 | Yes              | Read Committed, Serializable                                 | MVCC, Calvin                                           |
+| TiDB                 | Distributed DBMS | Yes          | Yes                 | Yes              | Read Committed, Repeatable Read                              | MVCC, Percolator                                       |
+| Oceanbase            | Distributed DBMS | Yes          | Yes                 | Yes              | Read Committed, Repeatable Read, Serializable                | MVCC, 2PC (Two-Phase Commit)                           |
+
+
+### DBMS Detailed Overview
+
+#### MySQL
+
+MySQL is a widely used open-source relational database management system suitable for applications of various scales, including web and enterprise solutions. It supports transaction processing and replication, offering good performance and stability. Transactions in MySQL are implemented via the InnoDB storage engine, which supports four isolation levels (Read Uncommitted, Read Committed, Repeatable Read, Serializable) and manages concurrent transactions using Multi-Version Concurrency Control (MVCC) and row-level locks.
+
+#### PostgreSQL
+
+PostgreSQL is a powerful open-source relational database system supporting complex queries and transactions, along with advanced features like JSON support, full-text search, and geospatial queries. It is widely used in enterprise and research applications. PostgreSQL handles transactions via MVCC, supports four isolation levels, and manages concurrency with shared and exclusive locks.
+
+#### MariaDB
+
+MariaDB is a community-driven open-source database management system created by the founders of MySQL. It is compatible with MySQL and includes enhanced features and performance optimizations. It is suited for applications requiring high performance and open-source access. Transaction management in MariaDB is similar to MySQL, primarily implemented via the InnoDB engine.
+
+#### Microsoft SQL Server
+
+SQL Server, developed by Microsoft, is a relational database management system designed for large-scale enterprise applications and data-driven solutions. It offers high availability, security, and enterprise-level support, with broad development tool and platform support. SQL Server supports five isolation levels and manages concurrency using locks and lock escalation. It also supports Snapshot Isolation, utilizing optimistic concurrency control mechanisms.
+
+#### Oracle Database
+
+Oracle Database is a powerful commercial-grade relational database management system used for enterprise data management and applications. It features high scalability, security, and data integrity. Oracle manages transactions through read consistency and locking mechanisms, supporting two main isolation levels (Read Committed, Serializable). It uses Multi-Version Concurrency Control (MVCC) for consistent reads and combines it with Two-Phase Locking (2PL).
+
+#### MyRocks
+
+MyRocks is a high-performance persistent storage engine developed by Facebook, designed to handle large-scale OLTP workloads. Based on RocksDB, it focuses on high throughput and low latency. Transactions in MyRocks are implemented using MVCC and key-based locks, supporting two isolation levels (Read Committed, Repeatable Read).
+
+#### MongoDB
+
+MongoDB is a document-oriented NoSQL database suitable for handling large amounts of unstructured data and applications requiring a flexible data model, such as web apps, analytics, and real-time data processing. MongoDB supports transactions starting from version 3.0, with isolation levels mainly being Read Uncommitted and Read Committed. It uses multi-document transactions and distributed locks for concurrency management and employs optimistic concurrency control (OCC).
+
+#### Redis
+
+Redis is an in-memory data structure store often used as a cache and message broker. It supports various data structures (e.g., strings, hashes, lists) and offers fast access and high performance. It is suited for real-time data processing and session storage. Redis does not support traditional transaction isolation levels due to its single-threaded model, relying on optimistic locks and transaction commands (MULTI/EXEC) for simple transaction support.
+
+#### Cassandra
+
+Cassandra is a distributed column-family NoSQL database known for its high availability and scalability. It is suitable for applications requiring large-scale data processing and distributed data storage, such as analytics and log management. Cassandra does not support traditional transaction isolation levels but manages data consistency through consistency levels (e.g., One, Quorum, All). It uses optimistic concurrency control and lightweight transactions (LWT) and employs a Timestamp Ordering Protocol (T/O).
+
+#### Neo4j
+
+Neo4j is a graph database management system focusing on storing and processing graph data (nodes and edges). It is suitable for applications requiring complex relationship and network analysis, such as social networks, recommendation systems, and network security analysis. Neo4j manages transactions using ACID properties, with the main isolation level being Read Committed, and employs locks and MVCC for concurrency control.
+
+#### InfluxDB
+
+InfluxDB is an open-source database designed specifically for time-series data, used for handling large volumes of timestamped data, such as monitoring, IoT sensor data, and real-time analytics. It offers high performance and query efficiency. InfluxDB does not support traditional transaction isolation levels as it focuses on time-series data's write and query performance, using the uniqueness of time-series data to prevent conflicts. InfluxDB employs optimistic concurrency control (OCC).
+
+#### TimescaleDB
+
+TimescaleDB is an open-source time-series database built on PostgreSQL, combining SQL and NoSQL features for high-performance analysis and querying of time-series data. It is suitable for time-series data storage and complex queries. TimescaleDB inherits PostgreSQL's transaction model, supports four isolation levels, and uses MVCC for concurrency management.
+
+#### CockroachDB
+
+CockroachDB is a distributed relational database management system with transaction support and high availability, designed for large-scale distributed applications requiring strong consistency and horizontal scalability, such as online services and global deployments. CockroachDB handles transactions via distributed transaction protocols (e.g., Percolator), supports two isolation levels (Read Committed, Serializable), and uses optimistic concurrency control and deterministic concurrency control protocols (Calvin).
+
+#### TiDB
+
+TiDB is an open-source distributed relational database management system compatible with MySQL protocol, supporting transaction processing and horizontal scalability. It is suitable for mixed workloads involving Online Transaction Processing (OLTP) and Online Analytical Processing (OLAP). TiDB's transactions are managed using Percolator, supporting two isolation levels (Read Committed, Repeatable Read).

src/dbtest/src/doc/Database_Research_Doc_zh.md

@@ -0,0 +1,65 @@
+### DBMS 大致介绍
+
+| DBMS 名称            | DBMS 类型              | 是否提供 ODBC | 是否支持事务 | 是否被 3TS-Coo 验证 | 支持的事务隔离级别                                           | 并发控制算法及事务描述       |
+| -------------------- | ---------------------- | ------------- | ------------ | ------------------- | ------------------------------------------------------------ | ---------------------------- |
+| MySQL                | 关系型数据库管理系统   | 是            | 是           | 是                  | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC，乐观和悲观控制         |
+| PostgreSQL           | 关系型数据库管理系统   | 是            | 是           | 是                  | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC                         |
+| MariaDB              | 关系型数据库管理系统   | 是            | 是           | 是                  | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC                         |
+| Microsoft SQL Server | 关系型数据库管理系统   | 是            | 是           | 是                  | Read Uncommitted, Read Committed, Repeatable Read, Serializable, Snapshot | 2PL，MVCC，乐观和悲观控制    |
+| Oracle Database      | 关系型数据库管理系统   | 是            | 是           | 是                  | Read Committed, Serializable                                 | 2PL (无等待、等待死锁)，MVCC |
+| MyRocks              | 关系型数据库管理系统   | 是            | 是           | 是                  | Read Committed, Repeatable Read                              | MVCC                         |
+| MongoDB              | NoSQL 数据库管理系统   | 是            | 3.0 后支持   | 是                  | Read Uncommitted, Snapshot                                   | 乐观控制（OCC）              |
+| Redis                | NoSQL 数据库管理系统   | 否            | 否           | 否                  | 不适用                                                       | 乐观锁，MULTI/EXEC           |
+| Cassandra            | NoSQL 数据库管理系统   | 是            | 是           | 是                  | 不适用（通过一致性级别控制）                                 | LWT，时间戳排序协议（T/O）   |
+| Neo4j                | 图数据库               | 是            | 是           | 否                  | Read Committed                                               | MVCC                         |
+| InfluxDB             | 时间序列数据库管理系统 | 是            | 是           | 否                  | 不适用                                                       | 乐观并发控制（OCC）          |
+| TimescaleDB          | 时间序列数据库管理系统 | 是            | 是           | 否                  | Read Uncommitted, Read Committed, Repeatable Read, Serializable | MVCC                         |
+| CockroachDB          | 分布式数据库管理系统   | 是            | 是           | 是                  | Read Committed, Serializable                                 | MVCC，Calvin                 |
+| TiDB                 | 分布式数据库管理系统   | 是            | 是           | 是                  | Read Committed, Repeatable Read                              | MVCC，Percolator             |
+| Oceanbase            | 分布式数据库管理系统   | 是            | 是           | 是                  | Read Committed, Repeatable Read，Serializable                | MVCC，2PC（两阶段提交）      |
+
+
+
+### DBMS 详细介绍
+
+#### MySQL
+MySQL 是一个广泛使用的开源关系型数据库管理系统，适合用于各种规模的应用，包括 Web 应用和企业级解决方案。支持事务处理和复制，具有良好的性能和稳定性。MySQL 的事务通过 InnoDB 存储引擎实现，支持四种隔离级别（Read Uncommitted, Read Committed, Repeatable Read, Serializable），并使用多版本并发控制（MVCC）和行级锁来管理并发事务。
+
+#### PostgreSQL
+PostgreSQL 是一个强大的开源关系型数据库系统，支持复杂的查询和事务处理，以及高级功能如 JSON 支持、全文搜索和地理空间查询。广泛用于各种企业和科研领域的应用。PostgreSQL 的事务通过 MVCC 实现，支持四种隔离级别，并使用共享锁和排他锁来管理并发事务。
+
+#### MariaDB
+MariaDB 是一个社区驱动的开源数据库管理系统，由 MySQL 的创始人创建，兼容 MySQL，并包含增强的功能和性能优化。适合需要高性能和开放源代码的应用。MariaDB 的事务管理与 MySQL 类似，也主要通过 InnoDB 引擎实现。
+
+#### Microsoft SQL Server
+SQL Server 是由 Microsoft 开发的关系型数据库管理系统，适用于大型企业应用和数据驱动的解决方案。提供高可用性、安全性和企业级支持，支持广泛的开发工具和平台。SQL Server 支持五种事务隔离级别，并使用锁和锁升级来管理并发事务。SQL Server 还支持快照隔离（Snapshot Isolation），使用乐观并发控制机制。
+
+#### Oracle Database
+Oracle Database 是一种功能强大的商业级关系数据库管理系统，用于企业级数据管理和应用程序。具有高度可扩展性、安全性和数据完整性。Oracle 的事务通过读一致性和锁来管理，并支持两种主要的隔离级别（Read Committed, Serializable）。Oracle 使用多版本并发控制（MVCC）来实现一致性读，并结合两阶段锁定协议（2PL）。
+
+#### MyRocks
+MyRocks 是 Facebook 开发的高性能持久性存储引擎，设计用于处理大规模的 OLTP 工作负载。基于 RocksDB，专注于高吞吐和低延迟的需求。MyRocks 的事务通过 MVCC 和基于键的锁实现，并支持两种隔离级别（Read Committed, Repeatable Read）。
+
+#### MongoDB
+MongoDB 是一个面向文档的 NoSQL 数据库，适合处理大量非结构化数据和需要灵活数据模型的应用，如 Web 应用、分析和实时数据处理。MongoDB 从 3.0 版本开始支持事务，隔离级别主要为 Read Uncommitted 和 Read Committed。MongoDB 使用多文档事务和分布式锁来管理并发事务，采用乐观并发控制（OCC）。
+
+#### Redis
+Redis 是一个内存中的数据结构存储，通常用作缓存和消息代理。支持多种数据结构（如字符串、哈希表、列表等），提供快速访问和高性能。适用于实时数据处理和会话存储等场景。Redis 不支持事务隔离级别，因为它是一个单线程处理模型，使用乐观锁和事务命令（MULTI/EXEC）来实现简单的事务支持。
+
+#### Cassandra
+Cassandra 是一个分布式的面向列的 NoSQL 数据库，具有高可用性和可伸缩性。适用于需要处理大量数据和分布式数据存储的应用，如分析和日志管理。Cassandra 不支持传统的事务隔离级别，而是通过一致性级别（如 One, Quorum, All）来控制数据一致性。Cassandra 使用乐观并发控制和轻量级事务（LWT）来管理并发事务，同时使用时间戳排序协议（T/O）。
+
+#### Neo4j
+Neo4j 是一个图数据库管理系统，专注于存储和处理图形结构数据（节点和边）。适用于需要复杂关系和网络分析的应用，如社交网络、推荐系统和网络安全分析。Neo4j 的事务管理使用 ACID 特性，主要隔离级别为 Read Committed，并使用锁机制和 MVCC 来管理并发事务。
+
+#### InfluxDB
+InfluxDB 是一个专为时间序列数据设计的开源数据库，用于处理大规模的时间戳数据，如监控、IoT 传感器数据和实时分析。具有高性能和查询效率。InfluxDB 不支持传统的事务隔离级别，因为它专注于时间序列数据的写入和查询性能，通过时间序列的唯一性来避免数据冲突。InfluxDB 使用乐观并发控制（OCC）。
+
+#### TimescaleDB
+TimescaleDB 是建立在 PostgreSQL 之上的开源时间序列数据库，结合了 SQL 和 NoSQL 的特性，用于处理时间序列数据的高性能分析和查询。适合于时间序列数据存储和复杂查询。TimescaleDB 继承了 PostgreSQL 的事务模型，支持四种隔离级别，并使用 MVCC 来管理并发事务。
+
+#### CockroachDB
+CockroachDB 是一个分布式的关系数据库管理系统，具有事务支持和高可用性，适用于需要强一致性和水平扩展的大规模分布式应用，如在线服务和全球部署的应用。CockroachDB 的事务通过分布式事务协议（如 Percolator）实现，支持两种隔离级别（Read Committed, Serializable），并使用乐观并发控制和确定性并发控制协议（Calvin）。
+
+#### TiDB
+TiDB 是一个开源的分布式关系数据库管理系统，兼容 MySQL 协议，支持事务处理和横向扩展。适用于在线事务处理（OLTP）和在线分析处理（OLAP）的混合工作负载。TiDB 的事务通过 Percolator 实现，支持两种隔离级别（Read Committed, Repeat)。