Earlier this week, Oracle announced the availability of MySQL Autopilot, the latest addition to MySQL Heatwave, Oracle’s in-memory query accelerator offered in the Oracle Cloud Infrastructure (OCI).
Oracle announced the availability of MySQL Autopilot
This blog will describe some of the features of this new offering and why enterprises and technology providers should consider using it.
Heatwave: More Insights, Much Faster
MySQL Heatwave, announced by Oracle in late 2020, is a parallel in-memory query accelerator that runs with Oracle MySQL as a service in OCI- dramatically improving the performance of queries by orders of magnitude.
This enables users of Heatwave to run both Online Transaction Processing (OLTP) and Online Analytical Processing (OLAP) workloads directly against a single MySQL database without the need for ETL tools and a separate analytics database instance.
Heatwave mirrors the transactions from the MySQL OLTP database in the OCI Object datastore, which makes these transactions available immediately for analytics queries without the delays inherent with a traditional ETL process.
Autopilot: Hidden AI-powered DBA
Oracle Autopilot is a new component of the Heatwave service that uses machine learning to automate the configuration and management of Heatwave clusters.
Autopilot provides a lot of the guidance and assistance that a DBA would normally provide – recommending optimal architectural configurations, monitoring and optimizing queries, deciding which tables will benefit from being partitioned in-memory, monitoring clusters, and even automatically creating new nodes and rebalancing workloads in a Heatwave cluster to replace any nodes that stop communicating.
Autopilot will provide guidance and recommendations for the number of nodes to provision in a Heatwave cluster based on analysis of the tables in a MySQL schema. It provides the ability to easily select all or some of the tables in the schema to include in the in-memory store, auto-provisions the cluster based on your selection, and then optimizes the operation of the Heatwave cluster.
How Does Heatwave With Autopilot Stack Up to the Competition?
Oracle’s Heatwave with Autopilot cluster can scale to 64 nodes with 32 TB RAM while maintaining a scalability of 0.89 across all nodes.
According to benchmarks provided in the Oracle announcement, Oracle’s MySQL Heatwave with Autopilot significantly outperforms the other leading Cloud-based MySQL offerings such as AWS, Snowflake, Google, and Microsoft- ranging from 3 to 9 times faster with consistently lower costs.
Oracle significantly outperformed its competitors in both OLTP and OLAP workloads, even when competing with a database tailored to support one or the other.
Bottom Line
Oracle appears to have significantly upped the ante in the Cloud DB market. Based on the benchmarks reported by Oracle, the Oracle MySQL Heatwave Autopilot solution is setting a new bar for Cloud database performance standards and is also simplifying the stack for those applications that have both OLTP and OLAP workloads. This also bodes well for OCI as this will be a performance and cost savings draw for some technology providers to move their application workloads to OCI.