OpenLDAP

Historically the OpenLDAP server (slapd, the Standalone LDAP Daemon) architecture was split between a frontend that handles network access and protocol processing, and a backend that deals strictly with data storage. This split design was a feature of the original University of Michigan code written in 1996[9] and carried on in all subsequent OpenLDAP releases. The original code included one main database backend and two experimental/demo backends. The architecture is modular and many different backends are now available for interfacing to other technologies, not just traditional databases.

Note: In older (1.x) releases, the terms "backend" and "database" were often used interchangeably. To be precise, a "backend" is a class of storage interface, and a "database" is an instance of a backend. The slapd server can use arbitrarily many backends at once, and can have arbitrarily many instances of each backend (i.e., arbitrarily many databases) active at once.[10]

Currently 17 different backends are provided in the OpenLDAP distribution, and various third parties are known to maintain other backends independently. The standard backends are loosely organized into three different categories:

• Data storage backends – these actually store data
  • back-bdb: the first transactional backend for OpenLDAP, built on Berkeley DB, removed with OpenLDAP 2.5.[11]
  • back-hdb: a variant of back-bdb that is fully hierarchical and supports subtree renames, removed with OpenLDAP 2.5.[12]
  • back-ldif: built on plain text LDIF files[10]
  • back-mdb: a transactional backend built on OpenLDAP's Lightning Memory-Mapped Database (LMDB)[10]
  • back-ndb: a transactional backend built on MySQL's NDB cluster engine, removed with OpenLDAP 2.6.[13]
  • back-wiredtiger: an experimental transactional backend built on WiredTiger, introduced with OpenLDAP 2.5.[10]
• Proxy backends – these act as gateways to other data storage systems
  • back-asyncmeta: an asynchronous proxy with meta-directory features, introduced with OpenLDAP 2.5.[10]
  • back-ldap: simple proxy to other LDAP servers[10]
  • back-meta: proxy with meta-directory features[10]
  • back-passwd: uses a Unix system's passwd and group data[10]
  • back-relay: internally redirects to other slapd backends[10]
  • back-sql: talks to arbitrary SQL databases, deprecated with OpenLDAP 2.5.[10]
• Dynamic backends – these generate data on the fly
  • back-config: slapd configuration via LDAP[10]
  • back-dnssrv: Locates LDAP servers via DNS[10]
  • back-monitor: slapd statistics via LDAP[10]
  • back-null: a sink/no-op backend, analogous to Unix /dev/null[10]
  • back-perl: invokes arbitrary perl modules in response to LDAP requests, deprecated with OpenLDAP 2.5.[10]
  • back-shell: invokes shell scripts for LDAP requests, removed with OpenLDAP 2.5.[14]
  • back-sock: forwards LDAP requests over IPC to arbitrary daemons[10]

Some backends available in older OpenLDAP releases have been retired from use, most notably back-ldbm which was inherited from the original UMich code, and back-tcl which was similar to back-perl and back-shell.[15]

Support for other backends will soon be withdrawn as well. back-ndb is removed now since the partnership with MySQL that led to its development was terminated by Oracle after Oracle acquired MySQL. back-bdb and back-hdb have been removed in favor of back-mdb since back-mdb is superior in all aspects of performance, reliability, and manageability.

In practice, backends like -perl and -sock allow interfacing to any arbitrary programming language, thus providing limitless capabilities for customization and expansion. In effect the slapd server becomes an RPC engine with a compact, well-defined and ubiquitous API.

Ordinarily an LDAP request is received by the frontend, decoded, and then passed to a backend for processing. When the backend completes a request, it returns a result to the frontend, which then sends the result to the LDAP client. An overlay is a piece of code that can be inserted between the frontend and the backend. It is thus able to intercept requests and trigger other actions on them before the backend receives them, and it can also likewise act on the backend's results before they reach the frontend. Overlays have complete access to the slapd internal APIs, and so can invoke anything the frontend or other backends could perform. Multiple overlays can be used at once, forming a stack of modules between the frontend and the backend.

Overlays provide a simple means to augment the functionality of a database without requiring that an entirely new backend be written, and allow new functionalities to be added in compact, easily debuggable and maintainable modules. Since the introduction of the overlay feature in OpenLDAP 2.2 many new overlays have been contributed from the OpenLDAP community.

Currently there are 25 overlays in the core OpenLDAP distribution, with another 24 overlays in the user-contributed code section, and more awaiting approval for inclusion.[16]

• Native slapd modules
  • acl/posixgroup – support posixGroup membership in access controls[17]
  • comp_match – support component-based matching[17]
  • kinit – maintain/refresh a Kerberos TGT for slapd[17]
  • passwd/ – additional password hashing mechanisms. Currently includes Kerberos, Netscape, RADIUS, and SHA-2.[17]
• SLAPI plugins
  • addrdnvalue – add RDN value to an entry if it was omitted in an Add request[19]

The basic synchronization operation is described in RFC 4533.[23] The protocol is defined such that a persistent database of changes is not required. Rather the set of changes is implied via change sequence number (CSN) information stored in each entry and optimized via an optional session log which is particularly useful to track recent deletes. The model of operation is that a replication client (consumer) sends a "content synchronizing search" to a replication server (provider). The consumer can provide a cookie in this search (especially when it has been in sync with the provider previously). In the OpenLDAP implementation of the RFC 4533, this cookie includes the latest CSN that has been received from the provider (called the contextCSN).

The provider then returns as search results (or, see optimization below, sync info replies) the present (unchanged entry only used in the present phase of the refresh stage) (no attributes), added, modified (represented in the refresh phase as an add with all current attributes), or deleted (no attributes) entries to put the consumer into a synchronized state based on what is known via their cookie. If the cookie is absent or indicates that the consumer is totally out of sync, then the provider will, in the refresh stage, send an add for each entry it has. In the ideal case, the refresh stage of the response contains only a delete phase with just a small set of adds (including those that represent the current result of modifies) and deletes that have occurred since the time the consumer last synchronized with the provider. However, due to limited session log state (also non-persistent) kept in the provider, a present phase may be required, particularly including the presentation of all unchanged entries as a means (inefficient) of implying what has been deleted in the provider since the consumer last synchronized.

The search can be done in either refresh or refreshAndPersist mode, which implies what stages occur. The refresh stage always occurs first. During the refresh stage, two phases may occur: present and delete, where present always occurs before delete. The phases are delimited via a sync info response that specifies which phase is completed. The refresh and persist stages are also delimited via such sync info response. An optional optimization to more compactly represent a group of entries that are to be presented or deleted is to use a sync info response containing a syncIdSet that identifies the list of entryUUID values of those entries.

The present phase is differentiated from the delete phase as follows. Entries that present unchanged entries may only be returned in the present phase. Entries that delete entries may only be provided in the delete phase. In either phase, add entries (including adds that represent all current attributes of modified entries) can be returned. At the end of a present phase, each entry that the consumer has that was not identified in an add entry or present response during the present phase is implicitly no longer in the provider and thus must be deleted at the consumer so as to synchronize the consumer with the provider.

Once the persist stage begins, the provider sends search results that indicate only the add, modify and delete of entries (no present unchanged entry indications) for those entries changed since the refresh stage completed. The persist stage continues indefinitely, meaning that search has no final "done" response. By contrast, in the refresh mode only a refresh stage occurs and such stage completes with a done response that also ends the present or delete phase (whichever phase was currently active).

This protocol keeps a persistent database of write accesses (changes) and can represent each modify precisely (meaning only the attributes that have changed). It is still built on the standard syncrepl specification, which always sends changes as complete entries. But in delta-syncrepl, the transmitted entries are actually sent from a log database, where each change in the main database is recorded as a log entry. The log entries are recorded using the LDAP Log Schema.[25]