Exadata Exascale Explained

Recently, the groundbreaking release of Oracle Exadata Exascale has sparked great excitement among tech enthusiasts, myself included. In the upcoming series of blogs, we will delve into the deployment and technical intricacies of this innovative platform. But before we embark on that journey, let's address two fundamental questions: What exactly is Exadata Exascale, and what benefits does it offer?

What is Exadata Exascale?

So, what sets Exadata Exascale apart? It boasts a revolutionary loosely coupled architecture that leverages Remote Direct Memory Access (RDMA) and SQL offload to streamline data management. By implementing a direct I/O path architecture, Oracle Database 23ai can now directly communicate with storage, eliminating the need for intermediary storage management tiers like ASM. This results in accelerated communication with storage nodes, paving the way for enhanced performance and efficiency. Exascale's direct I/O architecture plays a pivotal role in unlocking the potential of Exadata RDMA Memory (XRMEM), a unique memory cache in Exadata storage. With Exascale's ability to intelligently shift data between disk, memory, and flash, it delivers a performance blend of DRAM, flash IOPS, and disk capacity. Currently available in select regions, Exadata Database Service on Exascale Infrastructure is set to expand its footprint globally in the near future.

What are the benefits of Exadata Exascale?

Moving on to the benefits of Exadata Exascale, it revolutionises database storage and virtual machine utilisation by seamlessly integrating cloud attributes such as hyper-elasticity, multi-tenancy, and resource pooling. This means that accessing the power of Exadata is now more cost-effective and accessible than ever before.

Before Exascale, deploying an Oracle Database on Exadata infrastructure (ExaCS) required provisioning 2 compute nodes and 3 storage servers with a minimum of 4 OCPUs, resulting in a substantial upfront cost. In contrast, the Exadata Database Service on Exascale Infrastructure (ExaDB-XS) offers a more economical entry point. By starting with a cluster of 2 VMs equipped with 8 ECPUs, 22 GB of memory, and 300 GB of Exadata database storage, users can enjoy significant cost savings of up to 70.57%. Moreover, ExaDB-XS allows for flexible scaling of VMs and storage capacity, catering to evolving business needs with ease.

• 10-VM Cluster
• 200 Total ECPUs per VM
• 2 TB filesystem storage per VM
• 100 TB Exascale Vault database storage per VM Cluster

• Oracle Database 23.4
• Grid Infrastructure 23.4
• Exadata Image 24.0

Additionally, Exascale simplifies the process of creating database or PDB clones for development and testing purposes. With its redirect-on-write technology, clones can share blocks with the parent until modifications are made, reducing storage requirements and enhancing performance for development, testing, and recovery tasks. This streamlined approach empowers developers to efficiently create numerous PDB clones for their teams, facilitating agile development processes.

Conclusion

In conclusion, the ExaDB-XS service offers a cost-effective solution for deploying new workloads, enabling organisations to efficiently scale and manage their operational capabilities. With its seamless scalability and robust development cloud storage efficiency, ExaDB-XS bridges the worlds of cloud and Exadata, providing a comprehensive solution for expanding businesses.

If you'd like to discuss Exadata Exascale further, please don't hesitate to contact us, alternatively, check out our full range of Oracle Engineered Systems Services.

About the Author

René Antúnez, Principal Solutions Architect at Eclipsys Solutions, is an Oracle ACE Pro with a background in Oracle projects, including Exadata, ODA, and EM 12c implementation to name a few. Outside of work, he enjoys movies, music, and sharing the best of Mexico with the world through photography. Explore his personal blog for more insightful content, where you can discover a library of his blogs and social profiles to engage with.