Oracle today is following through on its announced plans to make its MySQL HeatWave database management systems available on the Amazon Web Services Inc. cloud.

MySQL HeatWave is a major upgrade to the popular open-source database engine that combines online transaction processing, analytics, machine learning and machine learning-based automation within a single MySQL instance. Oracle said the new features, which were introduced two years ago, eliminate the need for extract/transfer/load duplication between separate databases such as Amazon Aurora for transaction processing and Amazon Redshift for analytics.

Oracle said the new offering is optimized for AWS and delivers superior economics over services from Amazon and others. It cited a 4-terabyte TPC-H benchmark that it said showed MySQL HeatWave performing seven times better than Redshift and 10 times better than Snowflake Inc.’s cloud data warehouse. It also said its machine learning extensions perform 25 times better than Redshift ML.

MySQL HeatWave on AWS provides millisecond-level latencies for applications and a full-featured console that facilitates schema and data management and executes queries interactively. MySQL Autopilot, a component that uses machine learning techniques to automate many HeatWave features, is included with the console. Its features include automated provisioning, parallel loading, encoding, data placement, scheduling, query plan improvement, change propagation and error handling.

Available everywhere

Although Oracle and AWS compete in the public cloud, Oracle has always said HeatWave would be available on other cloud platforms. AWS customers that want to use the platform face “several challenges, including exorbitant data egress fees charged by AWS, the high latency of accessing the database from applications running on AWS, and the need to integrate with other applications running in AWS,” said Nipun Agarwal, senior vice president of research and advanced development at Oracle.

HeatWave is a core part of what Oracle calls a distributed cloud strategy that will make the engine available on multiple clouds as well as on-premises as part of Oracle Dedicated Region Cloud@Customer. “Customers can also replicate data from their on-premises MySQL OLTP applications to MySQL HeatWave on AWS or Oracle Cloud Infrastructure to obtain near real-time analytics,” Agarwal said. “MySQL HeatWave is always running the latest version of the MySQL database, which is not the case for many of the other MySQL-based services.”

Oracle also said the offering doesn’t compete with its own Autonomous Database. The two “address very different markets,” Agarwal said. “MySQL HeatWave supports popular open-source web applications such as WordPress, Magento and Drupal, whereas Autonomous Database supports Oracle Fusion Cloud SaaS applications. HeatWave is currently targeted at open-source cloud database users and developers with databases of less than 50 terabytes [while] Autonomous Database can scale to practically unlimited levels.”

Features added

MySQL HeatWave has also been beefed up with new security features that include server-side data masking, de-identification, asymmetric data encryption and a database firewall. Asymmetric data encryption enables developers and database administrators to use digital signatures to confirm the identity of people signing documents, Oracle said. The database firewall provides real-time protection against database-specific attacks, such as SQL Injections.

HeatWave ML, which is an optional feature provided at no charge, provides in-database machine learning capabilities for training, inference and explanations to enable customers to securely use machine learning on real-time data without ETL, Oracle said. HeatWave ML automates the machine learning lifecycle and stores all trained models inside the MySQL database, eliminating the need to move them to a separate machine learning tool or service.

While the core MySQL engine is available under an open source license, “HeatWave is designed and optimized specifically for the cloud and as such, it is not open source,” Agarwal said.

Photo: Valeriya Zankovych