DBD: Data Replication System - Overview and Planning

1: Overview Of Data Replication

The installation and setup of Data Replication can be divided into several sequential steps. Each topic is addressed in a separate section in the remainder of this document. Note that the steps marked Optional are essentially just fine-tuning and can be addressed later after the simplest, default implementation of Data Replication is installed and running properly.

2: How Data Replication is defined

Data Replication is used to copy selected DB Distributor information to one or more different databases whenever such information is changed. Data is always "pushed" from DB Distributor to the other database(s). DB Distributor is considered to be the "master" copy of the data.

During normal DB Distributor activities, data updates are also written to the Replication Queue (actually another Providex table residing on the DB Distributor server). One or more Data Replication Server processes running continuously on the DB Distributor server machine will monitor the Replication Queue and copy the data to one or more specified databases.

If a database connection is temporarily unavailable, the data updates will queue up. When the connection is eventually restored, the updates are applied in the same order they were originally made. There is very little chance for errors to interrupt normal DB Distributor processing.

Advantages of implementing a Replication Queue:

  • Fewer resources are required per operator than would otherwise be required to write directly to the database.
  • Fewer licenses are required because there are fewer database connections.
  • Every operator's computer does not need to maintain a database connection.

Examples of occasions where it is useful to replicate DB Distributor data:

  • When data is to be structured for queries or reports by other data warehouse applications.
  • When data is needed for integration with DB Enterprise.
  • When sending inventory levels/order status to a front-end web application.
  • When sending customer data to a contact management system.
  • When sending order information to a production control system.
  • When sending general ledger postings to a corporate accounting system.
  • When data is used by DemandBridge Report Suite (DBRS) reports.
  • When data is used to create a real-time backup copy of DB Distributor.

Examples of specific DB Distributor tables that can be replicated:

  • AR1500 – Customer information
  • AR6500 – Open Invoices

To reduce the bandwidth needed to replicate data:

  • The entire DB Distributor table is copied only during installation/initialization.
  • Otherwise, only DB Distributor information that is updated or deleted will be sent.
  • Data is updated even if DB Distributor tables are manually changed.
  • Data is updated as soon as possible after the information is changed.

Other characteristics Data Replication:

  • The update happens no matter how the data is changed in DB Distributor.
  • The tools used in the Data Replication module operate at a low level within the Providex environment that DB Distributor uses.
  • No changes to existing DB Distributor programs are required to enable Data Replication.
  • No worry about programs that did not receive the changes.

Note: Data Replication is usually run, by default, as a single background process on the DB Distributor server to replicate a list of tables. But it is also possible to run multiple copies of the Data Replication server process.

Advantages are:

  • Improved performance.
  • Each queue server automatically numbers itself.
  • Each queue server has a separate log file.

3: Where Data Replication is used

Data Replication is typically used with DB Distributor in several main areas:

  • Data Warehouse
  • Interface to Other Applications
  • Disaster Recovery

Data Warehouse

A data warehouse is a copy of business data specifically structured for querying and reporting. Using Data Replication in a data warehouse:

  • Reduces the load on production systems.
  • Adds additional data that can be used in reports.

  • Uses tools that can work with specific databases or platforms.


    – Tool only works on NT, Windows Server, or only with Oracle

  • Adds performance tuning for reports.


    – Add indexes

    – Pre-process data

Interface to Other Applications

There are cases when there is a need to “push” information from DB Distributor to another application:

  • Data needed for integration with DB Enterprise.
  • Inventory levels/order status to front end web application.
  • Customer data to a contact management system.
  • Order information to a production control system.
  • General ledger postings to a corporate accounting system.

  • Information flow is one-way.


    – Data is pushed from DB Distributor to the other application

    – DB Distributor is considered to be the “master” copy of the information

Disaster Recovery

Data Replication can provide a real-time backup copy of DB Distributor data, whereas typical system backups have several disadvantages:

  • Backups are generally only done once a day, usually overnight.
  • As the workday progresses more and more information is unprotected until the next backup.
  • It takes time to restore a backup from tape or disk to a backup machine.

Data Replication has several advantages:

  • Almost all DB Distributor tables can be replicated.
  • Data is written to corresponding DB Distributor tables in the replicated database.
  • Data is as up-to-date as possible because data updated in DB Distributor is immediately pushed to the other database.
  • Only updated data is copied which reduces network traffic versus copying entire sets of DB Distributor files.

It is possible to provide a complete “hot-swap” DB Distributor system:

  • Backup system is loaded with all software needed to begin operation.
  • In case of disaster, users will login to the backup system and are ready to go.
  • Requires the discipline to ensure that any changes (outside of DB Distributor data) are also done on the backup machine.

Advantages of a hot-swap data copy:

  • Only need to pre-load the backup system with Providex and DB Distributor supplied Data Replication software.
  • In case of disaster, data can be copied from the backup system to a new DB Distributor system.
  • Advantage is the data is more up-to-date than a daily backup.
  • Simpler to administrate than a full hot-swap backup system.

Note: Data Replication should be considered as supplemental to a daily backup plan, not as a replacement. Any good disaster recovery plan should have more than one level of backup.

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.