存储系统的分类

2022-07-01 12:28:49

之前收集了一些存储产品，最近又重新整理了一下，对他们进行了简单的分类。每个对存储的分类可能不仅相同，我的分类完全按照自己的喜好来分，如和您的分类不同，仅供参考。只是做了搜集和分类，少量产品加了写介绍，希望以后有时间，加更多更详细的介绍。

[TOC]

1.存储引擎

1.1 Hash Table

1.1.1 dbm (database manager)

https://en.wikipedia.org/wiki/Dbm

The dbm library stores arbitrary data by use of a single key (a primary key) in fixed-size buckets and uses hashing techniques to enable fast retrieval of the data by key.

1.2 btree

1.2.1 berkerlydb

https://en.wikipedia.org/wiki/Berkeley_DB

http://baike.baidu.com/view/1281930.htm

Key/value数据模型

Berkeley DB最初开发的目的是以新的HASH访问算法来代替旧的hsearch函数和大量的dbm实现（如AT&T的dbm，Berkeley的 ndbm，GNU项目的gdbm）

oracle

Written in C, java

BTREE, HASH, QUEUE, RECNO storage

https://www.oracle.com/database/berkeley-db/db.html

http://www.oracle.com/technetwork/database/database-technologies/berkeleydb/overview/index.html

1.2.2 LMDB (Lightning Memory-Mapped Database)

LMDB is a Btree-based database management library modeled loosely on the BerkeleyDB API, but much simplified. The entire database is exposed in a memory map, and all data fetches return data directly from the mapped memory, so no malloc's or memcpy's occur during data fetches.

https://symas.com/products/lightning-memory-mapped-database/

https://github.com/LMDB/lmdb

http://www.lmdb.tech/doc/

1.2.3 BoltDB

Bolt is a pure Go key/value store inspired by Howard Chu's LMDB project.

https://github.com/boltdb

https://github.com/boltdb/bolt

1.3 LSM

1.3.1 LevelDB

https://github.com/google/leveldb

C++, google

1.3.2 RocksDB

http://rocksdb.org/

https://github.com/facebook/rocksdb/

C++/java

1.3.3GoLevelDB

https://github.com/syndtr/goleveldb

1.3.4gorocksdb

https://github.com/tecbot/gorocksdb

Go wrapper for RocksDB

1.3.5 levigo

Go wrapper for LevelDB

https://github.com/jmhodges/levigo

1.3.6 mongo-rocks

RocksDB Storage Engine Module for MongoDB

https://github.com/mongodb-partners/mongo-rocks

C++

1.4 LSH

1.4.1 bitcask

https://github.com/basho/bitcask

采用bitcask模型的有：beandb, Riak

erlang

日志结构的key/value存储系统Bitcas

http://blog.chinaunix.net/uid-20196318-id-154750.html

Bitcask存储模型

http://blog.csdn.net/qq910894904/article/details/37756377

1.5 FractalTree

1.5.1 PerconaFT

https://github.com/percona/PerconaFT

https://www.percona.com/doc/percona-server-for-mongodb/perconaft.html

https://github.com/percona/PerconaFT/wiki

Mysql存储引擎之TokuDB以及它的数据结构Fractal tree(分形树)

http://www.fxysw.com/thread-5061-1-1.html

https://en.wikipedia.org/wiki/Fractal_tree_index

TokuDB的索引结构–分形树的实现

http://www.cnblogs.com/chaosheng/p/5250037.html

1.6 dbm系列

1.6.1 QDBM (Quick DataBase Manager)

http://sourceforge.net/projects/qdbm/

http://qdbm.sourceforge.net/

http://qdbm.sourceforge.net/benchmark.pdf

used by nmdb

1.6.2 ndbm (New Database Manager)

https://en.wikipedia.org/wiki/Ndbm

http://infolab.stanford.edu/~ullman/dbsi/win98/ndbm.html

ndbm (standing for New Database Manager) is a Berkeley produced version from 1986 of the AT&T dbm database.

ndbm stores arbitrary data by use of a single key in fixed-size buckets.

1.6.3 SDBM (Substitute Database Manager)

Substitute Database Manager

https://github.com/davidar/sdbm

http://www.cse.yorku.ca/~oz/sdbm.bun

1.6.4 GDBM （GNU Database Manager）

GNU Database Manager

http://www.gnu.org.ua/software/gdbm/

1.6.5 tdb (Trivial Database)

Trivial Database

http://sourceforge.net/projects/tdb/

http://tdb.sourceforge.net/

1.6.6 CDB

http://cr.yp.to/cdb.html

1.6.7 TinyCDB

Tiny Constant Database

http://www.corpit.ru/mjt/tinycdb.html

1.7 双类型

1.7.1 Wiredtiger (btree, LSM)

C语言的

WiredTiger 本身也支持 LSM option (默认是 btree )

mangodb

https://github.com/wiredtiger/wiredtiger

http://www.wiredtiger.com/

http://source.wiredtiger.com/2.8.0/architecture.html#cache

1.7.2 Tokyo Cabinet and Kyoto Cabinet (B+tree,hash table)

http://baike.baidu.com/view/2640411.htm

http://fallabs.com/tokyocabinet/

https://en.wikipedia.org/wiki/Tokyo_Cabinet_and_Kyoto_Cabinet

Tokyo Cabinet and Kyoto Cabinet are two libraries of routines for managing key-value databases.

Kyoto Cabinet is the designated successor of Tokyo Cabinet

Tokyo Cabinet 是一个DBM的实现

Tokyo Cabinet features on-disk B+ trees and hash tables for key-value storage

Used by nmdb，Kyoto TreeDB

1.7.3 RaptorDB key value store (B+ 树或者 MurMur 哈希索引)

一个很小的、快速的嵌入式 NoSQL 存储模块，使用 B+ 树或者 MurMur 哈希索引

Implemented in .NET

http://raptordbkv.codeplex.com/

http://www.codeproject.com/Articles/190504/RaptorDB

http://www.codeproject.com/Articles/316816/RaptorDB-The-Key-Value-Store-V

1.8 SQL引擎类

1.8.1 InnoDB

http://www.oschina.net/p/innodb/

http://baike.baidu.com/view/1238935.htm

https://en.wikipedia.org/wiki/InnoDB

1.9 document类

1.9.1 RaptorDB document store

http://raptordb.codeplex.com/

http://www.codeproject.com/Articles/375413/RaptorDB-the-Document-Store

2 嵌入式

https://en.wikipedia.org/wiki/Embedded_database

2.1 SQLite

https://en.wikipedia.org/wiki/SQLite

2.2 UnQLite

盘点移动开发中最流行的5个数据库

http://www.evget.com/article/2014/11/21/21843.html

3 单机存储

3.1 单值KV存储

3.1.1 Memcache

3.1.2 nmdb

https://blitiri.com.ar/p/nmdb/

use qdbm, berkeley db, tokyo cabinet or tdb as database backends

C语言

3.1.3 Memcachedb

C语言,新浪

http://memcachedb.org/

a distributed key-value storage system designed for persistent
It conforms to memcache protocol
uses Berkeley DB as a storing backend

write 18868 w/s
read 44444 r/s

using replication for master/slave
6 policy for replication:

http://memcachedb.org/memcachedb-guide-1.0.pdf

3.1.4 Kyoto Tycoon

http://fallabs.com/kyototycoon/

C/C++，FAL Labs

Kyoto Tycoon is a lightweight database server with auto expiration mechanism, which is useful to handle cache data and persistent data of various applications. Kyoto Tycoon is also a package of network interface to the DBM called Kyoto Cabinet.

3.1.5 ThruDB

建立在Apache Thrift framework下的简单服务

支持多个数据存储后端，包括BerkeleyDB、Disk、MySQL,还拥有Memcache和Spread集成

http://code.google.com/p/thrudb/

Thrudb is a set of simple services built on top of the Apache Thrift framework that provides indexing and document storage services

3.2 结构化KV存储

3.2.1 Redis

3.2.2 ssdb

https://github.com/ideawu/ssdb

http://ssdb.io/zh_cn/

http://ssdb.io/docs/replication.html

http://ssdb.io/docs/config.html

https://github.com/ideawu/ssdb/blob/master/ssdb.conf

采用ssd，使用leveldb作为存储引擎，兼容redis接口

C/C++

3.2.3 ssdb-rocks

https://github.com/ideawu/ssdb-rocks

ssdb的另一个版本，采用ssd，使用rocksdb作为存储引擎，兼容redis接口

3.3.4 ardb

https://github.com/yinqiwen/ardb

redis-protocol compatible persistent nosql, it support multiple storage engines as backend like Google's LevelDB, Facebook's RocksDB, OpenLDAP's LMDB, WiredTiger, the default backend is Facebook's RocksDB.

C++

http://yinqiwen.github.io/ardb/2014/08/23/ardbintroduction/

3.2.5 (reborndb)QDB

兼容redis协议

Rocksdb and LevelDB

https://github.com/reborndb/qdb

3.2.6 Pika

Pika 的存储引擎, 基于Rocksdb 修改. 封装Hash, List, Set, Zset等数据结构

https://github.com/Qihoo360/pika

http://git.oschina.net/baotiao/pika

http://www.jianshu.com/p/d4f23120cbe4

首发丨360开源的类Redis存储系统:Pika

https://mp.weixin.qq.com/s?__biz=MzAwMDU1MTE1OQ==&mid=2653547160&idx=1&sn=befd195e2aa788775aaf1cc3b6f6fab3&scene=1&srcid=0512FqLKcLjVNH0zbKVlVBSO&key=b28b03434249256b1da1489d74564ea1a9d5b15207160026adc2a6ce0622b47c084d23b75e909dfe14f6173662cbdf5b&ascene=0&uin=MjMxMzM3NjIyMw%3D%3D&devicetype=iMac+MacBookPro12%2C1+OSX+OSX+10.11.4+build(15E65)&version=11020201&pass_ticket=WU1aAnx4aVkuID0Quq0HGuKNQB68CvjQzaTnnIhjJFZwLcPqGk1zilYX4uRvF9sd

首发丨360开源的类Redis存储系统:Pika

https://media.weibo.cn/article?id=2309403974295628970629

3.2.7 LedisDB

A high performance NoSQL like Redis powered by Go

LevelDB, goleveldb, LMDB, RocksDB, BoltDB or Memory

http://ledisdb.com/

https://github.com/siddontang/ledisdb

3.3 文档型

3.3.1 SisoDB

C#编写的，专门提供给SQL Server面向文档的db-provider

http://www.sisodb.com

3.4 SQL

3.4.1 MySQL

3.4.2 innostore

https://github.com/basho/innostore

Innostore is a simple Erlang API to Embedded InnoDB

4 单机存储的proxy集群方案

4.1 KV/Redis类

4.1.1 Twenproxy

http://www.oschina.net/p/twemproxy

https://github.com/twitter/twemproxy

静态的分布式Redis方案

基于dynamo的思想

Dynomite: NetFlix对dynamo的开源通用实现

http://www.infoq.com/cn/news/2014/11/dynomite-netflix-dynamo

Netflix open sources Dynomite to make any datastore distributed

https://gigaom.com/2014/11/03/netflix-open-sources-dynomite-to-make-any-datastore-distributed/

A generic dynamo implementation for different k-v storage engines
inspired by Dynamo whitepaper

4.1.5 dbcached

http://code.google.com/p/dbcached/

a distributed key-value memory caching system for QDBM or Berkeley DB base on Memcached and NMDB

4.2 SQL类

4.2.1 Mycat

http://www.mycat.org.cn/

MyCat：开源分布式数据库中间件
http://mp.weixin.qq.com/s?__biz=MzAwNjMxNjQzNA==&mid=208187004&idx=1&sn=60aba39c148711e95f00ec7ca2e13bb1&scene=0#rd

4.2.2 MySQL Fabric

4.2.3 TDDL

4.2.4 Cobar

4.2.5 Atlas

4.2.6 Heisenberg

4.2.7 Vitess

5 KV存储

5.1 riak

http://www.oschina.net/p/riak/

https://github.com/basho/riak

采用bitcask模型

Riak的实现是基于Amazon的Dynamo论文

erlang

Riak是以 Erlang 编写的一个高度可扩展的分布式数据存储，Riak的实现是基于Amazon的Dynamo论文，Riak的设计目标之一就是高可用。

http://docs.basho.com/riak/kv/2.2.3/

5.2 beandb

https://github.com/douban/beansdb

https://github.com/douban/beanseye

采用bitcask模型

distributed key-value storage system

took the ideas from Amazon's Dynamo

5.3 Project Voldemort

http://www.project-voldemort.com/voldemort/

Voldemort is a distributed key-value storage system

http://www.project-voldemort.com/voldemort/design.html

https://github.com/voldemort/voldemort

BDB-JE, MySQL, Read-Only

5.4 Scalaris

http://scalaris.zib.de/

分布式key value

erlang

supported the ACID properties for multi-key transactions

http://code.google.com/p/scalaris/

5.5 Aeospike

www.aerospike.com

http://www.aerospike.com/technologies/

http://www.aerospike.com/docs/

http://www.aerospike.com/docs/architecture/index.html

5.6 Tair

http://code.taobao.org/p/tair/

http://www.oschina.net/p/tair/

https://github.com/alibaba/tair

c/c++, 自研的mdb,fdb

5.7 dynomite

https://github.com/moonpolysoft/dynomite/wiki

http://www.oschina.net/p/dynomite/

6 文档型存储

6.1 MongoDB

6.2 CouchDB

CouchDB是文档型存储

http://couchdb.apache.org/

6.3 Membase / Couchbase

http://www.couchbase.com/

http://www.oschina.net/p/membase

http://www.oschina.net/p/couchbase-server

https://en.wikipedia.org/wiki/Couchbase_Server

http://www.couchbase.com/wiki/display/couchbase/Home

面向文档的 NoSQL 数据库管理系统

6.4 SequoiaDB

一个类mongodb的文档型存储
http://www.sequoiadb.com/cn/

6.5 RavenDB

http://ravendb.net/
a .net document database built on the Windows ESENT storage system

支持Linq,可以使用C#的Linq语法查询数据

6.6 OrientDB

虽然是文档型数据库，但是它的关系管理方式却和图形数据库相类似

http://www.orientechnologies.com/

http://orientdb.com/orientdb/

Distributed Graph Database with the flexibility of Documents

是兼具文挡数据库的灵活性和图形数据库管理链接能力的可深层次扩展的文档-图形数据库管理系统。可选无模式、全模式或混合模式下。支持许多高级特性，诸如ACID事务、快速索引，原生和SQL查询功能。可以JSON格式导入、导出文档。若不执行昂贵的JOIN操作的话，如同关系数据库可在几毫秒内可检索数以百记的链接文档图

Even if it is a document-based database, the relationships are managed as in graph databases with direct connections between records.

It supports schema-less, schema-full and schema-mixed modes.

6.7 RethinkDB

http://www.rethinkdb.com/

https://en.wikipedia.org/wiki/RethinkDB

http://baike.baidu.com/link?url=IaJEn3OOWg3i_q4HSIyqBfxRUATc2bpsoNfzcHdyAVa7vbs02R6PsIIyR4F2V4zZy9Wq3FJYW550B-d6IF6MV_

RethinkDB 1.14 (Brazil) 发布，分布式数据库
http://www.oschina.net/news/54771/rethinkdb-1-14-brazil

7 列式存储

7.1 HBase

7.2 Cassandra

7.3 Accumulo

https://accumulo.apache.org/
Apache Accumulo is based on Google's BigTable design and is built on top of Apache Hadoop, Zookeeper, and Thrift.

7.4 Hypertable

http://hypertable.org/
an open source database system inspired by publications on the design of Google's BigTable.
Hypertable runs on top of a distributed file system such as the Apache HDFS, GlusterFS or the Kosmos File System (KFS).