LDAP Directory Technology Review

Perhaps the most important advantage that makes directory servers distinguished is the ability of each entity in the schema to have multiple instances of some attributes, something which is repeated by the literature reviewed on the field of directory servers.

Directory Servers enable the development of more complex architectures that would not be possible via traditional RDBMSs. The ability to hold multiple instances of attributes for a single entry is especially useful for deploying the proposed system and can not be matched using the relational model.

The flexibility available when designing the directory schema simplifies the design process, as most of the technologies required to complete this project already offer support for it, by providing ready made schemas and simplify the integration process. The relationships between entities are described in the directory schema and do not have to follow the strict rules of the relational model.

Additionally, because the proposed system needs to provide customised access to parts of the data stored within the directory and for various groups of users, the authentication and inheritance nature of the Directory Servers can provide the necessary security control mechanisms without compromising the integrity of the system.

LDAP directories and specifically OpenLDAP use minimal resources to perform data retrieval operations, something which is a major advantage over the traditional RDBMS system used by web hosting companies, which is MySQL. This reduces the load on the server during busy times.

Storing data in a directory and sharing it amongst applications saves you time and money by keeping administration effort and system resources down, (Steven Tuttle, 2004). Given that each application requires precious network resources to operate (E.g.: CPU and memory), Directory Servers can cope with even bigger demand.

Directory Servers can be queried for data “up to ten times faster than SQL database systems” (Henning Brauer, 2004), which provides obvious advantages to the system.

Directory Servers can not support real-time updating of data, as this requires the schema to be re-loaded and all the indexes defined rebuilt. In an environment with a large volume of updates, Directories are not suitable because of the server resource consumption required to do those updates.

Because of the distributed nature of the proposed system, the Directory Server can hold configuration and other data required to manage multiple systems. For the proposed system, OpenLDAP will be used to store account and network information which will be accessed by the network services (E.g.: Apache, Qmail, etc.).

Entries are organized in a distinct hierarchy (Brian Arkills, 2002). The hierarchical “classification” of entities and attributes within the directory enable Directory Developers with the power to define a custom authentication and privilege inheritance system which will provide the necessary security controls that would secure the data provided by the directory, without the need to make schema alterations.

Additionally, because each of the products and services offered by the individual departments require different information from the customers, developing such a solution using a RDBMS would require heavy programming and data integrity could be compromised.