ArangoDB

class: center, middle

![Alt text](https://www.arangodb.com/wp-content/themes/triagens/images/logos/arangodb_logo2.png)

---

# Agenda

1. Introduction
2. ArangoDB Basics
3. Advanced Features
4. ArangoDB vs MongoDB
5. Conclusion
6. Q & A

---

# Introduction

## Key features:

* multi-model nosql database
  * documents
  * graphs
  * key-values
* transactions
* SQL-like query (AQL)

```javascript
// ArangoDB.explain()
{
  "type": "NoSQL database",
  "openSource": true,
  "version": 2.4.3,
  "builtWith": ["C", "C++", "js"],
  "javascript": ["client side", "server side"],
  "uses": "v8"
}
```

---

# ArangoDB Basics

.left-column[
  ### Documents and collections
]
.right-column[
* ArangoDB is a database that works with documents

* Documents are sets of name/value pairs (think of JSON objects), e.g.

```json
{ 
  "name": "Jan", 
  "age": 37 
}
```

* Documents are organised in collections (like tables in relational databases)
]

---

# ArangoDB Basics

.left-column[
  ### Documents and collections
  ### Data types: JSON
]
.right-column[
* ArangoDB operates on JSON documents

* JSON data can be saved as-is:

```javascript
db.users.save({
  "name": "Jan",
  "age": 37
});
```

]

---

# ArangoDB Basics

.left-column[
  ### Documents and collections
  ### Data types: JSON
  ### Schema-free
]
.right-column[

* Documents can be heterogenous - even in the same collection

e.g.

```json
{ "name": "Jan", "age": 37 }

{ "name": "Tim", "city": "Paris" }

{ "name": {
    "firstname": "Patrick",
    "lastname": "Star"
  },
  "active": true
}
```

]

---

# ArangoDB Basics

.left-column[
  ### Documents and collections
  ### Data types: JSON
  ### Schema-free
  ### Simple queries
]
.right-column[

* Query a document by its unique id/key:

```javascript
db._document('users/jan');
```

* Query by providing an example:

```javascript
db.users.byExample({ name: 'Jan' });
```

]

---

# ArangoDB Basics

.left-column[
  ### Documents and collections
  ### Data types: JSON
  ### Schema-free
  ### Simple queries
  ### Indexes
]
.right-column[

* ArangoDB allows querying on any document attributes or nested attributes

* For each collection there will always be a primary index which is a hash index for the document keys (`_key` attribute) 
of all documents in the collection

* Secondary indexes can be created for higher query performances:

* Hash indexes (equality queries)
  * Skiplists (range queries)
  * 2d geo indexes (location queries)

]

---

# Advanced Features

.left-column[
  ### References
]
.right-column[

* ArangoDB can store references between documents

* References are documents with two endpoints (document type: Edge)

* Reference documents can carry own attributes

]

---

# Advanced Features

.left-column[
  ### References
  ### References example
]
.right-column[

```javascript
// Users documents (document)
{
  "name": "Jan",
  "age": 37
}

{
  "name": "Tim",
  "city": "Paris"
}

// Relations documents (edge document)
{
  "_from": "users/jan",
  "_to": "users/tim",
  "type": "knows"
}

{
  "_from": "users/tim",
  "_to": "users/jan",
  "type": "knows"
}

```

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
]
.right-column[

A Graph consists of vertices and edges. Edges are stored as documents in edge collections. 
In general a vertex is stored in a document collection.

.content-image[

![Alt text](http://bnode.org/media/2010/08/rww_ont.gif)

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Vertices & Edges 
]
.right-column[

```javascript
// Start vertex
{
  "_id": "posts/1",
  "comment": [
    "great post",
    "something else"
  ],
  "organization": "n-fuse"
}
// End vertex
{
  "_id": "users/tim",
  "name": "Tim"
}
// Edge
{
  "_id": "created_by/1",
  "_from": "posts/1",
  "_to": "users/tim"
}

```

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Graph definition
]
.right-column[

The graph module adds an additional layer on top of these collections giving you the following guarantees:

* All modifications are executed transactional
* If you delete a vertex all edges will be deleted, you will never have loose ends
* If you insert an edge it is checked if the edge matches the definition, your edge collections will only contain valid edges

#### Warning

These guarantees are lost if you access the collections in any other way than the graph module!!!

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Graph example
]
.right-column[

```javascript
var module = require("org/arangodb/general-graph");
// Create an empty graph
var graph = module._create("post");
// Graph can have an arbitrary amount of vertex 
// collections, which are not part of 
// any edge definition of the graph
graph._addVertexCollection("posts");
graph._addVertexCollection("users");

// Create a graph with edge definitions 
// and orphan collections:
var rel = module._relation("created_by", [
  "posts"
], [
  "users"
]);
graph._extendEdgeDefinitions(rel);

// Show graph definition
graph;
[ Graph post EdgeDefinitions: [ 
  "created_by: [posts] -> [users]" 
] VertexCollections: [ ] ]
```

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Graph query
]
.right-column[

Graph functions: 
* graph._edges(examples)
* graph._vertices(examples)
* graph_query.toVertices(examples)
* graph_query.outEdges(examples)
* etc.

AQL query:

```SQL 
// Get all posts of n-fuse organization
// and their coresponding authors
FOR v IN GRAPH_TRAVERSAL("post", {
  organization: "n-fuse"
}, "outbound")
  RETURN v
```

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Transactions
]
.right-column[

Transactions in ArangoDB are atomic, consistent, isolated, and durable (ACID).

* Atomicity: either complete in their entirety or have no effect at all.

* Consistency: no constraints or other invariants will be violated during or after any transaction.

* Isolation: hide the modifications of a transaction from other transactions until the transaction commits.

* Durability: operations from transactions that have committed will be made persistent. 
The amount of transaction durability is configurable in ArangoDB, as is the durability on collection level.

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Transactions
  ### Transactions example
]
.right-column[

```javascript
// setup 
db._create('c1');
db._create('c2');

db._executeTransaction({
  collections: {
    write: [ 'c1', 'c1' ]
    // i.e. the write mode includes the read mode
    // read: [ 'c1' ]
  },
  action: function () {
    var db = require('internal').db;
    for (var i = 0; i < 100; ++i) {
      db.c1.save({ _key: 'key' + i });
      db.c2.save({ _key: 'key' + i });
    }
    db.c1.count(); // 100 
    // abort 
    throw 'something wrong!'
  }
});

```

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Transactions
  ### ArangoDB query language (AQL)
]
.right-column[

* ArangoDB can also answer complex SQL-like queries

* Such queries are expressed in ArangoDB's textual query language

* This language allows joins, sub-queries, aggregation etc.

]

---

# Advanced Features

.left-column[
  ### References
  ### Graph
  ### Transactions
  ### ArangoDB query language (AQL)
  ### AQL example
]
.right-column[

```sql 
For u IN users
  FILTER u.city == "Paris" && 
	 u.age >= 18
  LET followers = (
    FOR f IN users
      FILTER f.follow == u._id
      RETURN f
  )
  RETURN {
    "users": u,
    "popularity": LENGTH(followers)
  }
```

]

---

# MongoDB vs ArangoDB

.left-column[
  ### Comparison
]
.right-column[

.content-table[

<table style="width:100%">
<thead>
  <tr>
    <th></th>
    <th>ArangoDB</th>
    <th>MongoDB</th>
  </tr>
</thead>
<tbody>
  <tr>
    <td>Data model</td>
    <td align="center">
      <pre>Document-oriented<br/>Graph-oriented<br/>key-value</pre>
    </td>
    <td align="center">
      <pre>Document-oriented</pre>
    </td>
  </tr>
  <tr>
    <td>Query language</td>
    <td align="center">
      <pre>Javascript<br/>API calls<br/>AQL<br/>RESTful</pre>
    </td>
    <td align="center">
      <pre>Javascript<br/>API calls</pre>
    </td>
  </tr>
  <tr>
    <td>Data types</td>
    <td align="center">
      <pre>JSON</pre>
    </td>
    <td align="center">
      <pre>JSON<br/>BLOB</pre>
    </td>
  </tr>
  <tr>
    <td>Map reduce</td>
    <td align="center">No</td>
    <td align="center">Yes</td>
  </tr>
</tbody>
</table>

]

---

# MongoDB vs ArangoDB

.left-column[
  ### Comparison
]
.right-column[

.content-table[

<table style="width:100%">
<thead>
  <tr>
    <th></th>
    <th>ArangoDB</th>
    <th>MongoDB</th>
  </tr>
</thead>
<tbody>
  <tr>
    <td>Transactions</td>
    <td align="center">Yes</td>
    <td align="center">No</td>
  </tr>
  <tr>
    <td>Revision control</td>
    <td align="center">Yes</td>
    <td align="center">No</td>
  </tr>
  <tr>
    <td>Graph support</td>
    <td align="center">Yes</td>
    <td align="center">No</td>
  </tr>
  <tr>
    <td>Native driver</td>
    <td align="center">REST API</td>
    <td align="center">BSON</td>
  </tr>
</tbody>
</table>

]

---

# MongoDB vs ArangoDB

.left-column[
  ### Comparison
  ### The most important difference
]
.right-column[

Why we need a more expressive query language ?

.content-image[

![Alt text](http://www.sarahmei.com/blog/wp-content/uploads/2013/11/Screen-Shot-2013-11-09-at-8.29.40-PM.png)

]

---

# Conclusion

1. Introduction of document reference 
2. Complex query with AQL
3. Graph module for future use

---
class: center, middle

# Thank you, any questions?

![Alt text](http://mannerofspeaking.files.wordpress.com/2010/08/question.jpg)