Earlier this year, I started contributing to a social media centric project using React, Node, GraphQL, and ArangoDB....wait? what is that?

So what is ArangoDB?

ArangoDB is a highly available and scalable multi-model database to natively work with graphs, documents, and full text search in one database & query language.

ArangoDB is designed for fast development and easy scaling. The best part? It is opensource!!

What's a multi-model database?

A Multi-model database is a database that can store, index, and query data in more than one model. A model such as relational database, document-oriented database, graph database, key-value pairs.

For example, if your application requires a graph database and a key/value store (like redis), you would have to use second database technology to support it. Being multi-model, ArangoDB allows you to not only use one database for both but run ad-hoc queries on data stored in different models.

Hence, multi-model databases can provide an elegant solution to the challenge of managing heterogeneous data multi-model.

Here's how it compares to other databases

ArangoDB vs Neo4J, ArangoDB vs MongoDB

AQL is powerful

The ArangoDB Query Language (AQL) is a declarative language, meaning that a query expresses what result should be achieved but not how it should be achieved. It's quite human readable as well.

Here's a basic example with AQL

-- Inserting a document
INSERT {
    "name": "Karan",
    "role": "Full Stack Developer",
    "age": 21
} INTO Users

-- Reading documents
FOR user IN Users
    RETURN user

-- Reading documents with filter
FOR user IN Users
    FILTER user.name == "Karan"
    RETURN user

-- Reading specific document
RETURN DOCUMENT("Users", "<document-key>")

Read more about basic operations here

This is just scratching the surface of AQL. Some of the other great features are powerful graphql traversals, array operators, high-level functions for geo index based searches.

Example for a simple graph traversal

-- General syntax
[WITH vertexCollection1[, vertexCollection2]]

FOR vertex[, edge[, path]]
  IN [min[..max]]
  OUTBOUND|INBOUND|ANY startVertex
RETURN vertex[, edge[, path]]

-- Example with Users collection
WITH Users

FOR vertex, edge, path IN 1..1 OUTBOUND "users/document-key"
  RETURN vertex

ArangoDB Oasis

Oasis is a fully managed cloud offering by Arango which makes it easier than ever to scale your cluster. No more worrying about AWS EC2 instances!!

Personal experience

I found ArangoDB to be a delight, it's powerful graph traversals and flexibility of AQL helped with a lot of complicated features we were working on. Hopefully, many more people give this database a try.