Documentation Center

Content Manager system architecture

The system architecture consists of a core resting on databases, that communicates with external components through an integration layer. This architecture supports all major open standards, including XML, XSL, REST, WSDL, and W3C.

Overview

The following diagram provides an overview of all major components in the physical system architecture:

Content Manager core

The Content Manager core contains the Repository and handles the logic responsible for content creation, review, translation, and publication. The core includes the following features:

  • Supports an object-relational data model. This means that the table structure in the database remains the same even when clients define their custom metadata in the Repository.
  • Interacts directly with either Oracle or Microsoft SQL Server, as well as with the full-text indexing engine.
  • Provides specific connectivity functionality to link the system to external translation management systems and publishing engines.
  • Enables communication developed through .NET technology.
  • Includes several APIs:
    • A private OpenAPI for communication with Organize Space OpenAPI is based on the OpenAPI Specification As from Tridion Docs 15.1 (hotfix), the OpenAPI is public, but limited to interacting with Organize Space.
    • Private XAPIs that communicate with Collective Spaces
    • API 2.5, an assembly-level library
    • Additional APIs that communicate with Translation Builder, the crawler service and the background task service

Web Services API

With the Web Services API, customers can develop custom applications that integrate with Content Manager, such as tools to support their specific authoring, review, translation and publication processes. Web Services support three technologies: OpenAPI on the one hand, and the older WCF .SVC (Windows Communication Foundation) the older .ASMX, which enable remote use of the API25 library, on the other hand.

Web Services in both technologies (WCF .SVC and .ASMX) are deprecated and are now replaced by the OpenAPI.

Web applications

These web applications are accessed through Mozilla Firefox and Chromium-based browsers like Google Chrome or Microsoft Edge:
  • Organize Space — provides access the Repository and Content Manager configuration options.
  • Draft Space — an easy-to-use browser-based XML content editor.
  • Review Space — an easy-to-use content review application.

Client tools

Content Manager provides the following applications as desktop clients:
  • Publication Manager: a desktop client that allows technical writers and information architects to assemble modular content into deliverable publications.
  • Authoring Bridge: an add-in for major XML authoring tools; it allows technical writers to access the Repository through their XML authoring tool.
  • Condition Manager: a desktop client to define and manage conditions in the Repository.
  • Content Importer: a desktop client to import structured DITA content into Content Manager.

User authentication through Access Management

Tridion Access Management (or simplyAccess Management) provides implementors with a simplified approach to identity management and gives administrators a central location for ongoing management of access to applications. You can use the Tridion Docs Identity Provider or implement your choice of external identity provider. For more details, refer to the section on security and access management.

Customization

The Content Manager behavior can be customized through tools built on top of the Repository:
  • Plugins (such as the Write plugins or the Publish plugins).
  • Custom background tasks.
  • Custom APIs.

DITA Open Toolkit

All publish require a transformation of the XML content into the target format. When transforming to HTML, Content Manager uses the DITA-OT standards, either directly through DITA-OT use or using an engine that applies DITA-OT standards.

Stateless component model

All Content Manager components are developed according to the stateless component model. Following the stateless component model, a request made by a user is loaded into memory, fulfilled, and then unloaded. Subsequent requests by the same user are treated just like any request; the server makes no attempt to remember what the user has previously requested.

Stateless development brings two important advantages in terms of:

  • System availability and maintenance: the servers hosting the business and presentation logic can be swapped, provided their configurations are identical (for example, the database location should be the same on both servers). This means that servers can be brought down for system maintenance without any application downtime, since there is no single point of failure.
  • System performance: stateless components enable Network Load Balancing (NLB) and Component Load Balancing (CLB) to dispatch requests to the processor with the lightest workload.