Replicate Data Into Azure Database for MySQL

In today’s post I’ll talk about replicating data in Azure Database for MySQL. Data in replication allows you to synchronize data from MySQL Server running on prem, virtual machines or database services hosted by other cloud providers into the Azure Database for MySQL.

The data in replication is based on the binary (BIN) log file position-based native to MySQL. So, this is the same as if you were running it on-prem and running the BIN log replication for an enterprise-class database service.

The information in the BIN log is stored in different formats according to the database changes that are being recorded. Then sleeves are configured to read the binary log from the master and to execute the events in the BIN log on the sleeve’s local database. You would write the log on the primary server and then the sleeve knows where the primary is and pulls over that information and executes it on the secondary database.

Use cases

A best use case for data in replication is when you’re using a hybrid data solution. With the data in replication, you can keep data synchronized between your on premises servers and your database for MySQL, thus getting a cloud based secondary solution as a fall over replication, disaster recovery and business continuity.

When you want to have an application that is part cloud based and part local, the synchronization is useful for creating those hybrid applications. It’s also appealing when you have an existing local database server but want to move data to a region closer to end users if you’re geo-located.

Another common use case is multi-cloud synchronization, so for complex cloud solutions you can use the data in replication to synchronize data between Azure Database for MySQL and different cloud providers, including virtual machines and database servers hosted in the cloud. So, this is a great option for another layer of redundancy for those cloud deployments.

A few things to keep in mind:

    • The MySQL system database is not replicated, so if you make changes to accounts and permissions on the primary, they will not be replicated. These changes must be done manually on the replica server.
    • The server version must be 5.6 and later.
    • Primary and replica versions must match.
    • Each cable in the database must have a primary key for consistency.
    • Global transaction identifiers are not supported.

So, some things to consider but still a great option for the scalability, redundancy and business continuity.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.