Oracle MySQL Heatwave Lakehouse goes GA to query data

3 min read
Oracle MySQL Heatwave Lakehouse goes GA to query data

Head over to our on-demand library to view sessions from VB Transform 2023. Register Here

Oracle is formally getting into the data lakehouse business with the general availability of its MySQL Heatwave Lakehouse service today.

MySQL Heatwave is a managed database-as-a-service (DBaaS) offering that is built on top of the open source MySQL relational database platform that Oracle develops. The core MySQL database is designed to focus on Online Transaction Processing (OLTP) workloads. With Heatwave, it has been extended to also support Online Analytical Processing (OLAP). 

As with many relational databases, MySQL Heatwave typically is only able to query data directly stored within the database. The MySQL Heatwave Lakehouse changes that paradigm, enabling the database to query data that is stored in cloud object storage, commonly referred to as a data lake. The data lakehouse concept aims to bridge the gap between traditional databases and data warehouse technologies, which requires all data to be indexed and stored natively with the ease of use and low cost of a cloud data lake.

Oracle first previewed the MySQL Heatwave Lakehouse service in October 2022 and is now making the service generally available on Oracle Cloud Infrastructure (OCI) as well as Microsoft Azure. Oracle plans to make service available on Amazon Web Services later this year. The overall goal is to help enable even more usage of the service, regardless of where organizations have data, Oracle says.


VB Transform 2023 On-Demand

Did you miss a session from VB Transform 2023? Register to access the on-demand library for all of our featured sessions.


Register Now

“The performance is identical, whether the data is in the object store or in the database,” Nipun Agarwal, Oracle SVP of MySQL database and MySQL HeatWave told VentureBeat. “That gives users flexibility.”

How MySQL Heatwave Lakehouse works

MySQL Heatwave is designed not just to enable both OLTP and OLAP, but overall faster queries.

Agarwal explained that MySQL Heatwave is an in-memory query accelerator that takes data stored in the MySQL database and accelerates queries to provide analytics and data warehouse capabilities. That same in-memory acceleration is critical to enabling the lakehouse functionality.

Agarwal said the Oracle service allows customers to query data stored in object storage using MySQL. Organizations can upload their data in various commonly used file formats such as comma-separated values (CSV) as well in the Apache Parquet file format.

Of note, Oracle MySQL Heatwave does not currently support some of the popular open source data lake table formats, such as Apache Iceberg, which is widely supported by multiple vendors including Snowflake, Cloudera and even Databricks, which recently announced support alongside its own delta lake format. Agarwal noted that Oracle will expand to support other file formats in the future as customer demand dictates.

Data here, data there, data everywhere — MySQL Heatwave will query anywhere

Whether the data is locally stored in MySQL Heatwave or in a data lake, users query data using standard MySQL SQL queries, according to Agarwal. He emphasized that the actual processing is done by the MySQL Heatwave engine in-memory, while the data remains in object storage which avoids the need to make duplicate copies of data.

What’s also interesting, Agarwal noted, users won’t know what the source of the file is, whether it’s directly from the database or a data lake. Going a step further, it’s also possible to combine data from both native storage and data lake to execute queries.

“From the user’s perspective, it is going to be very seamless and transparent,” said Agarwal.

AI in MySQL Heatwave Lakehouse

Oracle overall has a number of ongoing efforts related to AI and generative AI in particular.

Last month Oracle founder Larry Ellison provided details on a generative AI service with Cohere, and Oracle has been positioning its cloud platform as a good place for vendors to build large language models (LLMs). 

On the database side, the MySQL Heatwave database benefits from Oracle’s AutoML capabilities that helps to enable the database for machine learning (ML) training workflows. There is not any specific generative AI functionality in Oracle MySQL Heatwave yet, but that could change in the future.

“From a big picture view, you can envision LLMs making their way into the breadth of the Oracle portfolio,” Steven Zivanic, Oracle global VP for database and autonomous services product marketing told VentureBeat.

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.

Source link