MariaDB Platform Capabilities


MariaDB Platform scales up with more cores, memory and storage and scales out with more database instances – making it easy to get started with a standalone instance, run on a bigger server, add replicas to scale reads and finally, to scale out reads, writes and storage with hundreds of instances and distributed SQL.

  • Compression

    Reduce the size of data on disk by 33-66% using InnoDB or MyRocks table compression, or compress only specific columns.

  • Partitioning

    Partition large tables by range, list or hash to improve query and index performance, and use multiple disks/different disks per partition.

  • Parallel query
    Parallel query

    Use multiple threads to execute aggregates, sorts, joins and more on multiple partitions at the same time – scale up linearly with cores.

  • Read replicas
    Read replicas

    Add read replicas using asynchronous and semi-sync replication, or use MaxScale as a replication server for hundreds.

  • Sharding

    Spread table partitions across multiple database instances to scale out reads, writes and storage transparently to applications.

Distributed SQL

MariaDB Platform supports distributed SQL with the Xpand storage engine, easily transforming a replicated or clustered database into a fully distributed database capable of scaling out to execute millions of transactions per second without sacrificing latency or standard SQL (e.g., joins). Alternatively, use both replicated and distributed tables by using Xpand for tables with lots of data and/or queries, and InnoDB or MyRocks for tables with much less.

High availability

MariaDB Platform includes two options to meet enterprise high availability requirements: replication with automatic failover, and multi-master clustering using synchronous replication and a shared-nothing architecture.

Automatic failover

MariaDB MaxScale provides MariaDB Platform with automatic failover, promoting an up-to-date replica if the primary becomes unreachable or unresponsive, and rerouting application queries to the new primary.

Multi-master clustering

MariaDB Enterprise Cluster provides MariaDB Platform with synchronous, multi-master clustering for mission-critical applications requiring continuous availability and strong consistency through infrastructure failures.

Zero-interruption failover

While most databases can meet standard high availability requirements with automatic failover and/or multi-master clustering, MariaDB Platform is the only database with advanced features comparable to Oracle Application Continuity. Together, they hide infrastructure and database failures from applications, resulting in zero-interruption failover.

  • Connection migration
    Connection migration

    MaxScale migrates backend database connections after performing an automatic failover – applications don’t have to create new connections because of a failover.

  • Session restore
    Session restore

    MaxScale restores database sessions on a newly promoted primary after it has performed an automatic failover – applications don’t have to restore sessions because of a failover.

  • Transaction replay
    Transaction replay

    MaxScale replays in-flight transactions after performing an automatic failover so they can continue – applications don’t have to retry transactions because of a failover.

Background Divider Element

Disaster recovery

MariaDB Platform includes a number of tools and features for implementing comprehensive disaster recovery plans, including non-blocking backups and point-in-time restore with MariaDB Enterprise Backup to online point-in-time rollback with MariaDB Flashback. In addition, delayed replicas (e.g., 1 hour, 12 hour and 1 day) can be used as part of an advanced strategy to prevent accidental or malicious data loss/corruption from reaching backups.

Enterprise Backup

Perform online, non-blocking backups (full or incremental), create cumulative backups and use them to restore databases to a specific point in time with MariaDB Enterprise Backup.


Undo the most recent transactions to recover from malicious or accidental data loss/corruption faster, and without incurring downtime, by using MariaDB Flashback to “rewind the database.”

Delayed replicas

With delayed replication, replicas can be staggered behind the primary in order to prevent data loss/corruption from reaching backups and recover data without needing a restore.

Background Divider Element


MariaDB Platform goes above and beyond standard security features such as roles and auditing, with advanced data protection and security layers such as complete, end-to-end encryption, a powerful database firewall to stop intruders and malicious attacks and dynamic data masking/obfuscation to protect customer data and prevent data breaches.

  • Encryption

    Encrypt all data in motion with secure connections (TLS), and at rest, including logs, with transparent data encryption (TDE).

  • Dynamic data masking
    Dynamic data masking

    Prevent sensitive and personally identifiable information from being exposed by masking query results before they are returned.

  • Database firewall
    Database firewall

    Protect data from malicious attackers by blocking queries based on their type, syntax, columns, time of day, user and more.

Oracle Database compatibility

MariaDB Platform is the only database solution with a 100% open source implementation of Oracle PL/SQL, making it easy for enterprises to migrate from Oracle Database without having to rewrite stored procedures, modify schemas and change application code. MariaDB customers have imported hundreds of thousands of lines of PL/SQL.

Data types and sequences

MariaDB Platform supports standard Oracle Database data types such as VARCHAR2, NUMBER, DATE, RAW, BLOB and CLOB as well as sequences using Oracle Database syntax (e.g., NEXTVAL and CURRVAL).

Stored procedures (PL/SQL)

MariaDB Platform supports standard Oracle Database packages and procedures written in PL/SQL, including cursors, loops, variables, exceptions and functions such as CAST, LENGTH and SUBSTR.


MariaDB Platform supports hybrid relational/JSON data models, allowing developers to extend relational data models with JSON documents and query them with standard SQL for faster development, or to query relational data as JSON documents, while at the same time ensuring DBAs can continue to enforce data integrity and correctness.

Schema flexibility

Add JSON columns to extend existing tables to provide developers with greater flexibility – speed up app development by allowing the data model to evolve on demand rather than requiring schema changes up front.

Standard SQL

Query and index JSON documents with standard SQL using a comprehensive set of JSON functions to read/write JSON documents, fields, array and objects – or return relational data as JSON documents.