Verify Data Replication with Compare and Repair

Compare and Repair for SymmetricDS Pro can compare two databases, report on the differences, and repair the data without interrupting uptime. It works cross-platform between any supported databases, uses parallel threads for performance, and minimizes network traffic.  Let’s look at how compare and repair works and why it is required by some businesses.

Find Data Differences

Using the Manage -> Compare screen in the SymmetricDS Pro web console, the user can request a comparison between two remote databases. Requests can run immediately or be scheduled for a date and time. The user chooses which tables to compare , how to detect changes, and how to report or repair them.  Differences can be detected quickly by row counts or more accurately using a checksum. The precision of comparison on decimals, floating points, and timestamps can be restricted  to work across different database platforms. For a system that is undergoing changes, in-flight data can be used so it’s not miscounted as differences.

Report and Repair Data

The reporting level can be selected for a quick report or  more detailed report.  The first level report lists each table and whether it matched or not.  The second level report counts rows that are matching, different, missing, or extra.  Finally, the most detailed report includes data differences as a SQL script with insert, update, and delete statements to repair the target database. Afterwards, the user can choose to have the repair script applied automatically to the target database.

How Data Verification Works

Compare and repair works across the existing HTTP/S links that SymmetricDS Pro uses for data replication. Therefore, it can compare remote databases across a wide-area network. To minimize network usage, it separates data into chunks and calculates a checksum that is sent back to the source. Similarly, to minimize database overhead, it orders and processes data outside of the database. Furthermore, parallel threads are used to process tables simultaneously for improved performance.

Why Do Data Differences Occur?

It’s possible for differences in data to occur for the following causes:

• Hardware or database experiences failure
• Replication is configured incorrectly
• Operational activity causes corruption by accident

Why Verify Data?

Here are some reasons businesses require verification of their data:

• Accuracy for critical business data
• Compliance with regulations, such as Sarbanes-Oxley Act (SOX)
• Audit of controls for certification, such as Service Organization Control Type 2 (SOC 2)

Use Data with Confidence

Compare and repair compliments database replication by verifying that data is a reliable source of information for insights, analytics, and intelligence. Incomplete or incorrect data can have real-world implications on a business. In healthcare, it could make a mistake with patient care. In retail, it could make a mistake in expansion. In finance, it could violate a sanctions list. Error-free data that has been verified is a good step towards data that it is useful and actionable.