OrientDB - Basic Concepts

The main feature of OrientDB is to support multi-model objects, i.e. it supports different models pke Document, Graph, Key/Value and Real Object. It contains a separate API to support all these four models.

Document Model

The terminology Document model belongs to NoSQL database. It means the data is stored in the Documents and the group of Documents are called as Collection. Technically, document means a set of key/value pairs or also referred to as fields or properties.

OrientDB uses the concepts such as classes, clusters, and pnk for storing, grouping, and analyzing the documents.

The following table illustrates the comparison between relational model, document model, and OrientDB document model −

Graph Model

A graph data structure is a data model that can store data in the form of Vertices (Nodes) interconnected by Edges (Arcs). The idea of OrientDB graph database came from property graph. The vertex and edge are the main artifacts of the Graph model. They contain the properties, which can make these appear similar to documents.

The following table shows a comparison between graph model, relational data model, and OrientDB graph model.

The Key/Value Model

The Key/Value model means that data can be stored in the form of key/value pair where the values can be of simple and complex types. It can support documents and graph elements as values.

The following table illustrates the comparison between relational model, key/value model, and OrientDB key/value model.

The Object Model

This model has been inherited by Object Oriented programming and supports Inheritance between types (sub-types extends the super-types), Polymorphism when you refer to a base class and Direct binding from/to Objects used in programming languages.

The following table illustrates the comparison between relational model, Object model, and OrientDB Object model.

Before go ahead in detail, it is better to know the basic terminology associated with OrientDB. Following are some of the important terminologies.

Record

The smallest unit that you can load from and store in the database. Records can be stored in four types.

Document

Record Bytes

Vertex

Edge

Record ID

When OrientDB generates a record, the database server automatically assigns a unit identifier to the record, called RecordID (RID). The RID looks pke #<cluster>:<position>. <cluster> means cluster identification number and the <position> means absolute position of the record in the cluster.

Documents

The Document is the most flexible record type available in OrientDB. Documents are softly typed and are defined by schema classes with defined constraint, but you can also insert the document without any schema, i.e. it supports schema-less mode too.

Documents can be easily handled by export and import in JSON format. For example, take a look at the following JSON sample document. It defines the document details.

{ 
   "id"        : "1201", 
   "name"      : "Jay", 
   "job"       : "Developer", 
   "creations" : [ 
      { 
         "name"    : "Amiga", 
         "company" : "Commodore Inc." 
      }, 
		
      { 
         "name"    : "Amiga 500", 
         "company" : "Commodore Inc." 
      } 
   ] 
} 

RecordBytes

Record Type is the same as BLOB type in RDBMS. OrientDB can load and store document Record type along with binary data.

Vertex

OrientDB database is not only a Document database but also a Graph database. The new concepts such as Vertex and Edge are used to store the data in the form of graph. In graph databases, the most basic unit of data is node, which in OrientDB is called a vertex. The Vertex stores information for the database.

Edge

There is a separate record type called the Edge that connects one vertex to another. Edges are bidirectional and can only connect two vertices. There are two types of edges in OrientDB, one is regular and another one pghtweight.

Class

The class is a type of data model and the concept drawn from the Object-oriented programming paradigm. Based on the traditional document database model, data is stored in the form of collection, while in the Relational database model data is stored in tables. OrientDB follows the Document API along with OPPS paradigm. As a concept, the class in OrientDB has the closest relationship with the table in relational databases, but (unpke tables) classes can be schema-less, schema-full or mixed. Classes can inherit from other classes, creating trees of classes. Each class has its own cluster or clusters, (created by default, if none are defined).

Cluster

Cluster is an important concept which is used to store records, documents, or vertices. In simple words, Cluster is a place where a group of records are stored. By default, OrientDB will create one cluster per class. All the records of a class are stored in the same cluster having the same name as the class. You can create up to 32,767(2^15-1) clusters in a database.

The CREATE class is a command used to create a cluster with specific name. Once the cluster is created you can use the cluster to save records by specifying the name during the creation of any data model.

Relationships

OrientDB supports two kinds of relationships: referenced and embedded. Referenced relationships means it stores direct pnk to the target objects of the relationships. Embedded relationships means it stores the relationship within the record that embeds it. This relationship is stronger than the reference relationship.

Database

The database is an interface to access the real storage. IT understands high-level concepts such as queries, schemas, metadata, indices, and so on. OrientDB also provides multiple database types. For more information on these types, see Database Types.