evitaDB

Visit evitadb.io for the full documentation, examples and guides.

evitaDB is a specialized database with easy-to-use API for e-commerce systems. It is a low-latency NoSQL in-memory engine that handles all the complex tasks that e-commerce systems have to deal with on a daily basis. evitaDB is expected to act as a fast secondary lookup/search index used by front stores.

We aim for an order of magnitude better latency (10x faster or better) for common e-commerce tasks than other SQL or NoSQL database solutions on the same hardware specification. evitaDB should not be used for storing and processing primary data.

Why should you consider using evitaDB instead of Elasticsearch, MongoDB or relational database?

evitaDB is a database specialized for e-commerce tasks and has everything you need to implement an e-commerce catalog
evitaDB is more performant than Elasticsearch or PostgreSQL on the same HW sizing in typical e-commerce scenarios
evitaDB has a ready to use API from the day one:
- GraphQL - targets rich JavaScript front-ends
- REST - targets server side applications
- gRPC - targets fast inter-server communication used in microservices architecture and is used for the evitaDB client drivers

What's current status of evitaDB?

evitaDB is currently under active development. evitaDB is supported by the company FG Forrest, which specializes in the development of e-commerce stores for large clients in the Czech Republic and abroad. evitaDB concepts have been proven to work well in production systems with annual sales exceeding 50 million €.

Engineers from FG Forrest cooperate with academic team from University of Hradec Králové, so our statements about evitaDB performance are backed by thorough (and unbiased) testing and research. All proofs can be found in this repository, and you can run tests on your HW to verify our conclusions.

What's the license of the evitaDB

evitaDB is licensed under the Business Source License 1.1. Technically, it is not an open source license, but is an open source friendly license, because it automatically converts to one after a period of time specified in the license.

We're fans of open source, and we've benefited a lot from open source software (even the database engine uses some of it). The database implementation has taken thousands of man-days and, if successful, will take several thousand more. We were lucky to get an EU grant that partially funded the initial implementation, but we need to build a self-sustaining project in the long run. Our company uses evitaDB for its own commercial projects, so the development of the database is guaranteed, but without additional income the development would be limited. That's why we have chosen this type of license, but in the end we allow you - our users - almost any use.

In a nutshell:

the BSL license covers a period of 4 years from the date of the software release
4 year old version of evitaDB becomes permissive Apache License, v.2
both BSL and Apache licenses allow you to use evitaDB for OSS and/or commercial projects free of charge
there is one exception - you may not offer and sell evitaDB as a service to third parties

That's it.

Read license FAQ

Prerequisities

To checkout Git repository on Windows you need to have long paths enabled:

git config --system core.longpaths true

evitaDB requires and is tested on OpenJDK 17.

Java applications support multiple platforms depending on the JRE/JDK vendor. All major hardware architectures (x86_64, ARM64) and operating systems (Linux, MacOS, Windows) are supported. Due to the size of our team, we regularly test evitaDB only on the Linux AMD64 platform (which you can also use on Windows thanks to the Windows Linux Subsystem). The performance can be worse, and you may experience minor problems when running evitaDB on other (non-Linux) environments. Please report any bugs you might encounter, and we'll try to fix them as soon as possible.

How to build evitaDB

evitaDB is built using Maven. You can build the entire project by running the following command in the root directory of the project:

mvn clean install

Maven setup

The build uses Maven toolchains to select the correct JDK version. You must have JDK 17 installed and configured in your in your Maven toolchains. You can find more information about Maven toolchains in the Maven Documentation.

In short, you need ~/.m2/toolchains.xml in your home directory next to ~/.m2/settings.xml:

<?xml version="1.0" encoding="UTF8"?>
<toolchains xmlns="http://maven.apache.org/POM/4.0.0"
            xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
            xsi:schemaLocation="http://maven.apache.org/TOOLCHAINS/1.1.0 https://maven.apache.org/xsd/toolchains-1.1.0.xsd">
  <toolchain>
    <type>jdk</type>
    <provides>
      <version>17</version>
      <vendor>openjdk</vendor>
      <id>jdk17</id>
    </provides>
    <configuration>
      <jdkHome>/path/to/your/jdk17/installation/directory</jdkHome>
    </configuration>
  </toolchain>
</toolchains>

How this repository is organized

documentation: research documents, documentation, specifications
evita_api: set of all supported data types in evitaDB, conversions to & from other types, common data structures, basic exception hierarchy
evita_db: Maven POM allowing to link all necessary libraries for embedded evitaDB usage scenario
evita_engine: implementation of the database engine
evita_external_api: web API implementation
- evita_external_api_core: shared logic for all web APIs, Undertow web server integration, annotation framework for APIs
- evita_external_api_graphql: implementation of GraphQL API
- evita_external_api_grpc: implementation of gRPC API
  - client: Java driver for client/server usage scenario
  - server: gRPC server
  - shared: shared classes between client & server (generated gRPC stubs)
- evita_external_api_rest: implementation of REST API
- evita_external_api_lab: GUI client and its API
evita_functional_tests: test suite verifying functional correctness of standard and edge cases of the API on a small amount of data, this library also contains unit tests for evita_db
evita_performance_tests: test suite executing most common operations on real world data that generates performance statistics of each implementation
evita_query: query language, query parser, utilities for query handling
evita_store: binary serialization using Kryo library, persistent key/value datastore implementation
evita_test_support: utility classes that make writing integration tests with evitaDB easier
jacoco: Maven POM that allows to aggregate test coverage for entire project

Quality requirements for the code

In order code to be accepted it will fulfill following criteria:

line coverage with unit tests will be >= 70%
all classes and methods will have comprehensible JavaDoc
there will be no TODO statements in the code
there will be no commented out code

Standing on giants' shoulders

evitaDB uses the following open source projects:

RoaringBitmap by Daniel Lemire, Richard Statin and others
Kryo by Nathan Sweet, Thomas Heigl, Martin Grotzke and others
Netty and Armeria by Trustin Lee, Norman Mauer, Ikhun Um and others
Logback by Ceki Gülcü and others
Jackson by Tatu Saloranta and others
gRPC Java by Eric Anderson, Carl Mastrangelo, ZHANG Dapeng, Kun Zhang and others
GraphQL Java by Andreas Marek, Brad Baker and others
Byte Buddy by Rafael Winterhalter and others

And many others. We are grateful for their work.

Icons sourced at Icons8.com

FgForrest/evitaDB