DBMS > Cachelot.io vs. Graph Engine vs. Milvus vs. Tarantool vs. TerarkDB

System Properties Comparison Cachelot.io vs. Graph Engine vs. Milvus vs. Tarantool vs. TerarkDB

Editorial information provided by DB-Engines

Name

Description

In-memory caching system

A distributed in-memory data processing engine, underpinned by a strongly-typed RAM store and a general distributed computation engine

A DBMS designed for efficient storage of vector data and vector similarity searches

In-memory computing platform with a flexible data schema for efficiently building high-performance applications

A key-value store forked from RocksDB with advanced compression algorithms. It can be used standalone or as a storage engine for MySQL and MongoDB

Primary database model

Key-value store

Graph DBMS
Key-value store

Vector DBMS

Document store
Key-value store
Relational DBMS

Key-value store

Secondary database models

Spatial DBMS

with Tarantool/GIS extension

DB-Engines Ranking measures the popularity of database management systems
Trend Chart

Score	0.00
Rank	#383	Overall
	#60	Key-value stores

Score	0.61
Rank	#240	Overall
	#21	Graph DBMS
	#35	Key-value stores

Score	2.31
Rank	#113	Overall
	#3	Vector DBMS

Score	1.72
Rank	#144	Overall
	#25	Document stores
	#25	Key-value stores
	#66	Relational DBMS

Score	0.00
Rank	#383	Overall
	#60	Key-value stores

Website

github.com/bytedance/terarkdb

Technical documentation

www.graphengine.io/docs/manual

milvus.io/docs/overview.md

www.tarantool.io/en/doc

bytedance.larkoffice.com/docs/doccnZmYFqHBm06BbvYgjsHHcKc

Developer

Microsoft

ByteDance, originally Terark

Initial release

2015

2010

2019

2008

2016

Current release

2.3.4, January 2024

2.10.0, May 2022

License

Commercial or Open Source

Open Source

Simplified BSD License

Open Source

MIT License

Open Source

Apache Version 2.0

Open Source

BSD-2, source-available extensions (modules), commercial licenses for Tarantool Enterprise

commercial

restricted open source version available

Cloud-based only

Only available as a cloud service

DBaaS offerings (sponsored links)

Database as a Service

Providers of DBaaS offerings, please contact us to be listed.

Zilliz Cloud – Cloud-native service for Milvus

Implementation language

C++

.NET and C

C++, Go

C and C++

C++

Server operating systems

FreeBSD
Linux
OS X

.NET

Linux
macOS

10.14 or later
Windows

with WSL 2 enabled

BSD
Linux
macOS

Data scheme

schema-free

yes

Flexible data schema: relational definition for tables with ability to store json-like documents in columns

schema-free

Typing

predefined data types such as float or date

yes

Vector, Numeric and String

string, double, decimal, uuid, integer, blob, boolean, datetime

XML support

Some form of processing data in XML format, e.g. support for XML data structures, and/or support for XPath, XQuery or XSLT.

Secondary indexes

yes

SQL

Support of SQL

Full-featured ANSI SQL support

APIs and other access methods

Memcached protocol

RESTful HTTP API

Open binary protocol

C++ API
Java API

Supported programming languages

.Net
C
C++
ColdFusion
Erlang
Java
Lisp
Lua
OCaml
OCaml
Perl
PHP
Python
Ruby

C#
C++
F#
Visual Basic

C++
Go
Java
JavaScript (Node.js)
Python

C
C#
C++
Erlang
Go
Java
JavaScript
Lua
Perl
PHP
Python
Rust

C++
Java

Server-side scripts

Stored procedures

yes

Lua, C and SQL stored procedures

Triggers

yes, before/after data modification events, on replication events, client session events

Partitioning methods

Methods for storing different data on different nodes

none

horizontal partitioning

Sharding

Sharding, partitioned with virtual buckets by user defined affinity key. Live resharding for scale up and scale down without maintenance downtime.

none

Replication methods

Methods for redundantly storing data on multiple nodes

none

Asynchronous replication with multi-master option
Configurable replication topology (full-mesh, chain, star)
Synchronous quorum replication (with Raft)

none

MapReduce

Offers an API for user-defined Map/Reduce methods

Consistency concepts

Methods to ensure consistency in a distributed system

none

Bounded Staleness
Eventual Consistency
Immediate Consistency
Session Consistency
Tunable Consistency

Casual consistency across sharding partitions
Eventual consistency within replicaset partition

when using asyncronous replication
Immediate Consistency within single instance
Sequential consistency including linearizable read within replicaset partition

when using Raft

Foreign keys

Referential integrity

yes

Transaction concepts

Support to ensure data integrity after non-atomic manipulations of data

ACID, with serializable isolation and linearizable read (within partition); Configurable MVCC (within partition); No cross-shard distributed transactions

Concurrency

Support for concurrent manipulation of data

yes

yes, cooperative multitasking

yes

Durability

Support for making data persistent

optional: either by committing a write-ahead log (WAL) to the local persistent storage or by dumping the memory to a persistent storage

yes

yes, write ahead logging

yes

In-memory capabilities

Is there an option to define some or all structures to be held in-memory only.

yes

yes, full featured in-memory storage engine with persistence

yes

User concepts

Access control

Role based access control and fine grained access rights

Access Control Lists
Mutual TLS authentication for Tarantol Enterprise
Password based authentication
Role-based access control (RBAC) and LDAP for Tarantol Enterprise
Users and Roles

More information provided by the system vendor

Specific characteristics

Milvus is an open-source and cloud-native vector database built for production-ready...
» more

Competitive advantages

Highly available, versatile, and robust with millisecond latency. Supports batch...
» more

Typical application scenarios

RAG: retrieval augmented generation Video media : video understanding, video deduplication....
» more

Key customers

Milvus is trusted by thousands of enterprises, including PayPal, eBay, IKEA, LINE,...
» more

Market metrics

As of January 2024, 25k+ GitHub stars 10M+ downloads and installations 3k+ enterprise...
» more

Licensing and pricing models

Milvus was released under the open-source Apache License 2.0 in October 2019. Fully-managed...
» more

We invite representatives of system vendors to contact us for updating and extending the system information,
and for displaying vendor-provided information such as key customers, competitive advantages and market metrics.

Related products and services

We invite representatives of vendors of related products to contact us for presenting information about their offerings here.

More resources

DB-Engines blog posts

Vector databases
2 June 2023, Matthias Gelbmann

show all

Data processing speed and reliability: in-memory synchronous replication
9 November 2021, Vladimir Perepelytsya, Tarantool (sponsor)

show all

Recent citations in the news

Trinity
2 June 2023, microsoft.com

Open source Microsoft Graph Engine takes on Neo4j
13 February 2017, InfoWorld

IBM releases Graph, a service that can outperform SQL databases
27 July 2016, GeekWire

Aerospike Is Now a Graph Database, Too
21 June 2023, Datanami

The graph analytics landscape 2019 - DataScienceCentral.com
27 February 2019, Data Science Central

provided by Google News

What Is Milvus Vector Database?
6 October 2023, The New Stack

Zilliz Unveils Game-Changing Features for Vector Search
22 March 2024, Datanami

Milvus 2.4 Unveils Game-Changing Features for Enhanced Vector Search
20 March 2024, GlobeNewswire

AI-Powered Search Engine With Milvus Vector Database on Vultr
31 January 2024, SitePoint

Zilliz Cloud boosts vector database performance
31 January 2024, InfoWorld

provided by Google News

TaranHouse: New Big Data Warehouse Announced by Tarantool
4 April 2018, Newswire

In-Memory Showdown: Redis vs. Tarantool
4 April 2023, Хабр

Deploying Tarantool Cartridge applications with zero effort (Part 1)
16 December 2019, Хабр

Deploying Tarantool Cartridge applications with zero effort (Part 2)
13 April 2020, Хабр

Тarantool Cartridge: Sharding Lua Backend in Three Lines
9 October 2019, Хабр

provided by Google News

Share this page