AWS Speeds Up Amazon Aurora Relational Database

Amazon Web Services Inc. (AWS), has increased the query time performance for its Aurora relational database service. This service was purpose-built to take advantage of cloud computing’s benefits.
Amazon Aurora is compatible with MySQL and PostgreSQL databases and is said to include the performance/availability of high-end commercial databases along with the simplicity and cost-effectiveness of open source offerings, while being faster than both.
The cloud-based fully managed RDBMS service is now faster thanks to new functionality that leverages cloud attributes such as scalability, distributed processing, and more.
AWS has announced that transactional data analysis queries can now be performed faster with Amazon Aurora Parallel Query.
AWS claims that faster analytical queries are possible without having to copy transactional data into separate systems. This is a new approach to parallelized queries. Amazon Parallel Query uses the processing power of all the nodes to speed up queries. Aurora data is spread across hundreds of storage nodes located in multiple locations within the product’s storage layer.
AWS stated that Parallel Query uses Aurora’s unique architecture to parallelize query processing across multiple CPUs. “While other databases can parallelize query process across CPUs in a few servers, Parallel Query utilizes Aurora’s unique architecture for pushing down and parallelizing query processing across thousands CPUs in the Aurora storage layers.” Parallel Query reduces network, CPU and buffer pool contention by offloading analytical query processing into the Aurora storage layer. This allows for the transactional workload to be shared with the Aurora storage layer.
Jeff Barr, AWS spokesperson, explained more in a blog posting. He provided a graphic to illustrate the Aurora storage nodes with data stored in fast SSDs across multiple shared storage volumes.
[Click on the image to see a larger view.] Barr stated that Amazon Aurora (source : AWS). “Each node of the storage layer pictured above also contains plenty of processing power.” “Aurora can now make great use of this processing power by taking your analytic queries (generally those that process large portions of a table) and running them in parallel across hundreds of thousands of storage nodes. This provides speed benefits that are close to two orders of magnitude. This new model reduces network, CPU and buffer pool contention so you can simultaneously run transactional and analytical queries on the same table, while still maintaining high throughput for both types.
Barr explains how to use the new feature in his post. He also mentions that although it is free, IO costs may rise due to its direct access storage.