MySQL HeatWave on OCI vs. Amazon Aurora
Accelerate MySQL performance by orders of magnitude for analytics and mixed workloads with Oracle MySQL HeatWave. According to a performance comparison with publicly available benchmarking code from Oracle, enabling anyone to run the benchmarks, MySQL HeatWave delivers 2800X better price performance than Amazon Aurora. MySQL HeatWave is the only service that enables customers to run OLTP and OLAP workloads directly from their MySQL database.
Customers choose MySQL HeatWave over Amazon Aurora for several reasons:
- Performance: MySQL HeatWave is 1400X faster than Amazon Aurora.
- Cost savings: MySQL HeatWave is half the cost of Amazon Aurora.
- Single database for transactional and analytics workloads: MySQL HeatWave eliminates the risk, cost, and complexity of using two separate databases.
- Machine learning-powered automation: MySQL AutoPilot uses advanced machine-learning techniques to automate HeatWave functions, making it easier to use and further improving performance and scalability.
- Existing applications benefit from real-time analytics without changes: No changes are required to existing applications to run OLTP and real-time analytics workloads simultaneously in a single MySQL database.
- Native, in-database machine learning: HeatWave AutoML provides the ability to build, train, deploy, and explain machine learning models within MySQL HeatWave, eliminating the need to move data to a separate machine learning service.
1. MySQL HeatWave is 1400X faster than Amazon Aurora
Perfected for performance and scalability, HeatWave is an innovative, in-memory query accelerator. MySQL HeatWave is faster than comparable cloud database services, including Amazon Aurora—at a fraction of the cost. This has been demonstrated by multiple standard industry benchmarks such as TPC-H, TPC-DS, CH-benCHmark—and on real-world customer workloads. Their analysis found that for a 4 TB TPC-H analytics workload, MySQL HeatWave is 1400X faster than Amazon Aurora.
See the performance details and learn more about the benchmark configuration
Capability and Evidence |
MySQL HeatWave |
Amazon Aurora |
|---|---|---|
Can customers run analytic and complex queries without indexing?HeatWave does not require indexes to accelerate queries. Data in the MySQL InnoDB storage engine is transparently transformed and propagated to HeatWave, becoming available for query acceleration as soon as it is updated in MySQL. This significantly reduces manual database tuning for database administrators (DBAs). Amazon Aurora requires DBAs to manually create and maintain indexes for better query performance. As more complex queries are run, more indexes are needed. Maintaining the indexes becomes tedious and time consuming for DBAs. For a 4 TB TPC-H dataset, it takes 5 days to create the needed indexes for query processing in Amazon Aurora, while it only takes 4 hours to load MySQL data into HeatWave; it is, therefore, available for analytics 30 times sooner. |
yes |
no |
Can customers run ad hoc analytic queries efficiently?HeatWave implements state-of-the-art algorithms for distributed in-memory analytic processing for queries with scans, joins, group-by, aggregation, and top-k. With MySQL AutoPilot – Auto Query Plan Improvement, HeatWave automatically optimizes query plans using statistics learned from the execution of previous queries. Ad hoc queries can run as fast as planned queries, eliminating manual tuning by DBAs. Without knowing the queries in advance, DBAs cannot create appropriate indexes in Amazon Aurora to improve query performance. |
yes |
no |
Can the database service scale as data grows and still maintain high performance?HeatWave is designed for massive scalability and performance. It has a highly partitioned architecture, which enables massive inter- and intra-node parallelism. Its intelligent query scheduler overlaps computation with network communication tasks to achieve very high scalability across thousands of cores. The performance improvements of MySQL HeatWave over Amazon Aurora increase with the data size. MySQL HeatWave is 151X faster than Aurora for 256 GB of data, 956X faster for 1 TB of data, and 1400X faster for 4 TB of data. While Amazon Aurora can scale out by adding replicas to handle concurrency, it can only scale up to increase the compute resources for query processing. |
yes |
no |
Red3i has indicated that migrating from Amazon Aurora to MySQL HeatWave has increased query performance for the company by as much as 1,000X for real-time analytics and by 85% for overall workloads, without changing code to its applications. IT costs dropped by 60% by eliminating multiple middleware tools and analytical platforms. Automatic recovery minimized downtime for high availability.
2. MySQL HeatWave is half the cost of Amazon Aurora
Compared to Amazon Aurora discounted partial upfront 1-year reserved pricing, MySQL HeatWave is half the cost and provides 2800X better price-performance based on a 4TB TPC-H data set.
Pablo Lemos, Cofounder and CTO of Tetris.co, describes how MySQL HeatWave dramatically reduced the company’s Amazon Aurora and Redshift costs by more than 50% while delivering real-time insights and enabling the company's expansion.
3. Single database for transactional and analytics workloads
Eliminate the risk, cost, and complexity of using two separate databases. Oracle MySQL HeatWave, with its integrated in-memory query accelerator, is the only service that enables database administrators and application developers to run OLTP and OLAP workloads directly from their MySQL database. For a 100 GB mixed workload with OLTP and OLAP queries, MySQL HeatWave is 18X faster, provides 110X better throughput than Amazon Aurora for OLAP queries, while maintaining the same performance as Aurora for OLTP queries.
Capability and Evidence |
MySQL HeatWave |
Amazon Aurora |
|---|---|---|
Can customers run hybrid OLTP and OLAP workloads efficiently using a single database service?MySQL HeatWave is the only MySQL service that enables customers to run both OLTP and OLAP workloads in MySQL– without the need to Extract, Transform, Load (ETL) data to a separate database for analytic processing. No changes to existing applications are necessary. Amazon Aurora’s parallel query provides only suboptimal performance improvement for OLAP queries. Amazon Aurora customers who want to run analytics workloads against their data need to use a separate OLAP database service—Amazon Redshift. Using two different databases plus the ETL process increases complexity and costs. Security and compliance risks also increase as data moves between data stores. |
yes |
no |
Can customers run real-time analytics workloads against their MySQL database?Changes made by MySQL transactions are propagated in real-time to HeatWave and become immediately available for analytics queries, enabling real-time analytics. Amazon Aurora users who need efficient analytics have no choice but to move their analytics workloads to a separate database to avoid lengthy delays from long running queries. By the time data is available in the separate analytics database, it’s already stale, so customers don’t get real-time analytics. |
yes |
no |
Can customers eliminate the cost, complexity, and risk of having to ETL by using a single database service?MySQL HeatWave provides a unified MySQL database platform for analytics and transactional workloads. This eliminates the need for the complex, time-consuming, expensive ETL and integration required by a separate analytics database. Amazon Web Services does not provide a single, unified service for OLTP and OLAP workloads. Using Amazon Aurora for OLTP workloads and Amazon Redshift for analytics requires customers to ETL all data to a different database, which increases complexity, risks, and costs. |
yes |
no |
For FANCOMI, migrating from Amazon Aurora to MySQL HeatWave not only increased performance by 10X to generate real-time analytics but also significantly reduced costs, since the company no longer had to move the data to an analytical database. Furthermore, FANCOMI did not have to modify its application to realize this dramatic performance improvement.
4. MySQL Autopilot provides machine learning-powered automation
MySQL Autopilot automates many of the most important and often challenging aspects of achieving high query performance at scale. MySQL Autopilot is available at no additional charge for MySQL HeatWave customers.
Capability and Evidence |
MySQL HeatWave |
Amazon Aurora |
|---|---|---|
Does the database service include built-in machine learning automation for operations?MySQL AutoPilot uses advanced machine learning to automate database lifecycle operations including provisioning, data loading, query processing, and error handling. Amazon Aurora lacks the equivalent built-in machine learning-powered automation, requiring DBAs to manually provision, maintain, and tune the database. |
yes |
no |
Does the database service include built-in machine learning automation for auto provisioning?MySQL Autopilot Auto Provisioning determines the number of HeatWave nodes required for running a workload by adaptive sampling of table data on which analytics is required. This means that customers no longer need to manually estimate the optimal size of their cluster. Amazon Aurora does not have the equivalent capability for auto provisioning. Developers need to guess or manually test the optimal instance type and cluster size for their workload, increasing cost and development time. |
yes |
no |
Does the database service include built-in machine learning automation for auto query plan improvement?MySQL Autopilot Auto Query Plan Improvement learns various statistics from the execution of queries and improves the execution plan of future queries. This improves the performance of the system as more queries are run. Amazon Aurora does not intelligently learn and improve query plans based on previously executed queries. DBAs need to periodically analyze and manually update table statistics for the Aurora optimizer to generate better query plans. |
yes |
no |
Does the database service include built-in machine learning automation for auto scheduling?MySQL Autopilot Auto Scheduling determines which queries in the queue are short running and prioritizes them over long running queries in an intelligent way to reduce overall wait time. Amazon Aurora lacks the equivalent machine learning-powered workload management capability. Queries are executed in a First In, First Out (FIFO) order, which can degrade query performance. |
yes |
no |
“For MySQL database developers, this is a game changer. Usually, developers scale MySQL through labor-intensive sharding schemes, often resulting in complex application SQL logic, and exposing the system to multiple levels of human error. With HeatWave, this is not an issue. MySQL developers also run into constant challenges configuring the database for maximum efficiency and coding the SQL for the best performance. With HeatWave, these are no longer concerns.”
—Carl Olofson, Research Vice President, Data Management Software, IDC
5. Existing applications benefit from real-time analytics without modifications
ISVs and enterprises want to accelerate their ability to make accurate predictions in order to improve business results. Customers running MySQL applications can immediately take advantage of HeatWave for real-time analytics, without making any changes to their applications.
Capability and Evidence |
MySQL HeatWave |
Amazon Aurora |
|---|---|---|
Can customers process transactions and get real-time analytics in the same database without making changes to existing applications?Existing MySQL applications work with HeatWave without any changes to benefit from real-time analytics. HeatWave is designed as a MySQL pluggable storage engine, which completely shields all the low-level implementation details from customers. As a result, they can manage both HeatWave and the MySQL database with the same management tools, including the OCI console, REST APIs, and command line interface. Amazon Aurora is not designed to provide real-time analytics. To efficiently analyze data, customers need to use a separate analytics database such as Amazon Redshift. | yes |
no |
"The MySQL HeatWave technology is by far the best in the market now...Wikibon strongly recommends that enterprise IT departments set a three-year plan to eliminate separate OLAP databases and ETL from MySQL transactional databases."
—David Floyer, CTO, Wikibon
6. MySQL HeatWave eliminates the need for a separate machine learning service
With HeatWave AutoML, customers can build, train, deploy, and explain machine learning models within MySQL HeatWave.
Capability and Evidence |
MySQL HeatWave |
Amazon Aurora |
|---|---|---|
Does the database service provide in-database machine learning?HeatWave AutoML enables developers and data analysts to build, train, deploy, and explain machine learning models within MySQL HeatWave.Amazon Aurora lacks the equivalent in-database machine learning capability. As a result, customers must move their data to a separate machine learning service (e.g., Amazon SageMaker) to build and train machine learning models. Security and compliance risks increase as data moves between systems. | yes |
no |
Are all machine learning models explainable, so users can understand and explain what happens in the models from input to output?HeatWave AutoML delivers predictions with an explanation of the results, improving regulatory compliance, fairness, repeatability, causality, and trust. All models created by HeatWave AutoML are explainable.All models in Aurora ML are not explainable, which can reduce trust in them, increase the risk of bias, and introduce complexity for regulatory compliance. | yes |
no |
Is the machine learning lifecycle automated?HeatWave AutoML fully automates the machine learning lifecycle, including algorithm selection, intelligent data sampling, feature selection, and hyperparameter tuning—saving customers significant time and effort. HeatWave AutoML enables developers and data analysts to build machine learning models using familiar SQL commands; no prior ML experience is required.Amazon Aurora ML does not automate many elements of the machine learning lifecycle and requires data science expertise to influence the performance, accuracy, and cost of the machine learning models’ training. | yes |
no |
Try MySQL HeatWave for free