As you probably know Exadata is an engineered system designed to host the entire Oracle software solution stack.
Reading about the Exadata architecture gets any of us database technologists excited. Seeing it in action makes you realize the product is not overhyped. What stands out to me every time I interact with the database is the extreme performance which is delivered by several unique features of the product.
Exadata storage provides database aware storage services, such as the ability to offload database processing from the database server to storage, and provides this while being transparent to SQL processing and database applications. Hence just the data requested by the application is returned rather than all the data in the queried tables.
Here we see a Quarter stack Exadata. The top two blocks represent the database. The bottom are the Exadata cells. The way these two units interact is part of Exadata’s secret sauce.
When using Exadata, much SQL processing is offloaded from the database server to the Exadata cells. Exadata enables function shipping from the database instance to the underlying storage in addition to providing traditional block serving services to the database. One of the unique things the Exadata storage does compared to traditional storage is return only the rows and columns that satisfy the database query rather than the entire table being queried. Exadata pushes SQL processing as close to the data (or disks) as possible and gets all the disks operating in parallel. This reduces CPU consumption on the database server, consumes much less bandwidth moving data between database servers and storage servers, and returns a query result set rather than entire tables. Eliminating data transfers and database server workload can greatly benefit data warehousing queries that traditionally become bandwidth and CPU constrained. Eliminating data transfers can also have a significant benefit on online transaction processing (OLTP) systems that often include large batch and report processing operations.