What constitutes Oracle Internet Directory & How does OID work?

An Oracle Internet Directory stores its data in an Oracle database; the necessary schemas

are in the pre-seeded database created as part of an Oracle Application Server infrastructure installation.

1. An Oracle Internet Directory instance is implemented as a single listening process that monitors a port for requests from LDAP clients, and a number of server processes that action the requests. These processes must all reside on the same machine, as the communication between them is via whatever IPC (interprocess communication) protocol is provided by the operating system.
   
2. The LDAP clients can be on any remote machine, since LDAP runs over TCP (though there may be firewall and security issues)
   
3. The database storing the information can also be on a remote machine, since the server processes connect to the database using Oracle Net.
   

The flow of information for an LDAP request over various protocols

1. A client process passes a search request to the listener process with
   
2. The listener process passes the request to a server process with
   
3. The server process passes the request to the database with Oracle
   
4. The database executes the query.
   
5. The database returns the result to the server with Oracle Net.
   
6. The server returns the result to the listener with IPC.
   
7. The listener returns the result to the client with LDAP

Oracle Internet Directory instances aren’t controlled by the administrator directly. It is done through a fly-by-wire approach implemented by a monitor process. The various control utilities (graphical and command-line) do not actually control the instances—they insert rows into a table in the database storing the directory.The monitor process then reads these commands and actions them.The monitor process is the oidmon process. The oidmon must be running before the oidldapd processes—the directory server processes—can be launched.

In some cases, one Internet directory is not sufficient. There may be a number of reasons for this, but the most obvious are scalability, network overhead, and singlepoint-of-failure.

The LDAP standard ensures that multiple directories contain identical copies of the directory tree through the concept of directory replication. Each directory runs an LDAP replication server, which propagates changes made to the entries in its local directory to the replication servers.

Similar in concept to directory replication is directory synchronization. The term replication means maintaining identical directory trees in different directories. Directory synchronization defines a mechanism whereby changes can be propagated between different directories. The objective is not to keep the directories identical, but merely to ensure that enough information is passed between them to eliminate the need for duplicating the data entry process.