Architecture

Overview

SAS MDM is built on a sophisticated data management platform that provides both development and run-time support for applications.

Development

During development, you can use DataFlux Data Management Studio (Data Management Studio) and DataFlux Data Management Server (Data Management Server). Data Management Studio provides the ability to design and test processes (batch jobs) and real-time services (data jobs), which can then be deployed and run using the following:
  • DataFlux Data Management Server
  • DataFlux Data Management Studio
  • scripting with a command line interface
The SAS MDM and SAS Data Quality software offerings rely on the Quality Knowledge Base (QKB) for data quality processing. The QKB provides capabilities such as standardization of data and computation of match codes. The QKB is used both during development and at run time.

Run Time

At run time, a multi-tier architecture that consists of presentation, business, and data tiers is used.
  • The presentation tier hosts the web interface in a Servlet 2.5-compliant container and includes a Java application server and a browser. Batch interfaces might be available for some operations such as bulk loads and updates.
  • The business tier runs on a Data Management Server and is implemented using Data Management Studio processes (batch jobs) and real-time services (data jobs). It implements a Service Oriented Architecture (SOA) that enables any system to use its services through SOAP requests. Common web service standards are used to maintain simplicity when integrating with other systems. Authentication within the business tier runs on a SAS Metadata Server. This server can authorize user permissions for applications, assign roles and security restrictions, and define authentication parameters for accessing other systems, such as the database, for each user or group. The real-time or transactional component of SAS MDM interfaces with other applications through an SOA environment. Specifically, the ability to call business services hosted on Data Management Server is accomplished through web services. Process jobs can also be used as business services that accept data on a record-by-record basis. This enables users to use SAS MDM data and services within an enterprise architecture.
  • The data tier hosts the supporting data and metadata for the system, which consists of a relational database such as Oracle or Microsoft SQL Server. The data model is documented and accessible to third-party tools and can also be accessed through Data Management Studio jobs.
The following display shows how the components fit together:
Architecture Diagram
Architecture Diagram

SAS MDM Components

Data Management Studio

Data Management Studio is a graphical user interface (GUI)-based application that provides access to other tools and services. You can use Data Management Studio for job flow management and customization, to manage data source connections, to manage user accounts through SAS Metadata Server, to deploy and execute jobs on the Data Management Server, and to run batch jobs and data services within a Windows environment.

DataFlux Process Jobs and Real-time Business Services

A DataFlux process job is a multi-step data management process that can be run from Data Management Studio, from Data Management Server, or from a command line interface or script. Real-time business services, also called data jobs, are deployed through Data Management Server and are used by the SAS MDM web application to provide data to the application through SOAP calls.

Quality Knowledge Base

The Quality Knowledge Base (QKB) contains the files, file relationships, and metadata needed to correctly parse, match, standardize, and otherwise process data as required for SAS MDM. The QKB can be customized to meet an organization's needs. The QKB can support different locales to support processing for many different languages.

Data Management Server

The Data Management Server (labeled as DM Server in Architecture Diagram) enables data management and integration processes to be deployed across an entire IT environment. SAS MDM uses batch jobs and real-time services deployed through Data Management Server to manage the master data hub and to provide data through the web interface and reports. Data Management Server handles all aspects of database connectivity and service availability through SOAP calls.
Note: In many instances, you should have at least two Data Management Servers working together through network load balancing or failover hardware or software, to achieve the desired results for performance and meet server uptime goals, particularly for real-time service processing.

Databases

SAS MDM relies on access to a Database Management System (DBMS, labeled as SAS MDM Schema in Architecture Diagram), such as Oracle or SQL Server, to act as a repository for metadata and data. The Master Data Management component of SAS MDM requires two database schemas for its operation.

SAS Data Management Console

The SAS Data Management Console is a central, web-based location for all master data management activities. This environment provides a launch location for all associated SAS MDM applications and SAS Visual Analytics reports, as well as other SAS applications.
The SAS MDM applications Master Data Manager, Data Remediation, and Task Manager enable you to perform the following functions:
  • Master Data Manager
    • create and manage a hub
    • define new entity types and their attributes
    • search for data within the hub
    • create and update data within the hub
  • Data Remediation
    • monitor and correct errors in data loaded into SAS MDM
  • Task Manager
    • start, stop, view, and interact with active workflow instances
  • Data Lineage (not available in all SAS MDM product bundles)
    • view relationships and perform impact analysis for selected objects stores in SAS metadata
  • Business Data Network (not available in all SAS MDM product bundles)
    • define and manage metadata for business terms
Other applications might appear in your SAS Data Management Console depending on the SAS MDM offering.
Several of these web applications use SAS data server repositories to store application data. These are represented in Architecture Diagram as schemas in the SAS Application and Metadata Tier.

SAS Workflow Studio

This application is a drag-and-drop workflow designer. Users can modify existing MDM workflow templates or create new ones for use with other SAS MDM processes. Workflows can also be uploaded and activated in the SAS server environment and can be downloaded for additional editing.

SAS Visual Analytics Applications

If your SAS MDM product bundle includes reporting capabilities, business users and data stewards have several reporting applications available to them to design, view, and share reports for SAS MDM and SAS Data Remediation. See the documentation for SAS Visual Analytics for more information about application capabilities.

Environment Manager

This application enables an administrator to see in one location key operational metrics of all SAS servers and other technologies deployed with SAS MDM. This application also includes server configuration and management capabilities.

SAS Web Server

This server handles web traffic and communication for all SAS web applications running in the mid-tier environment.

SAS Application Server

An application server hosts the SAS Data Management Console and other associated applications. This is installed and configured during the deployment process. No other third-party application servers are supported.

SAS Metadata Server

This server makes important system wide metadata available to SAS clients and other SAS servers. Interaction with this server is accomplished through SAS Management Console.

SAS LASR Analytic Server

If your SAS MDM product bundle includes reporting capabilities, then SAS LASR Analytic Server is present in your SAS MDM environment. You can load SAS MDM instance data to SAS LASR Analytic Server using SAS/ACCESS engines. This loads the data into memory making it available to SAS Visual Analytics applications.
Normal MDM functionality does not require an access engine. However, if your SAS MDM product bundle includes reporting capabilities delivered through SAS Visual Analytics, you must have the following installed on the computer that hosts SAS LASR Analytic Server:
  • for the SAS MDM reports, either SAS/ACCESS to Oracle if your SAS MDM target database is Oracle or SAS/ACCESS to ODBC if your SAS MDM target database is Microsoft SQL Server
  • for the SAS Data Remediation reports, SAS/ACCESS to Postgres

SAS Management Console

This application enables users to manage groups, users, and roles on the SAS Metadata Server for all SAS MDM web applications. It also is the location where various web application configuration parameters are set.
Last updated: April 19, 2017