Friday, October 16, 2020

Exadata Cloud Service X8M—No Database Workload is Too Large

DB Exam Prep, Database Tutorial and Material, Database Learning, Database Certification

The fastest cloud database service is now the most scalable


Oracle Exadata Cloud Service, the world’s fastest cloud database service, is now even faster, supporting the latest Exadata X8M hardware and bringing the benefits of Intel® Optane™ Persistent Memory (PMEM) and Remote Direct Memory Access over Converged Ethernet (RoCE) to the Oracle Cloud. On top of that, Exadata Cloud Service X8M redefines elasticity in the cloud, with configurations that scale to over 1,600 cores and 25 PB of HCC compressed data.

How did Exadata Cloud become even faster? With smart use of persistent memory and the RoCE networking fabric, Exadata Cloud now has 50x lower storage access latency than AWS databases. With typical storage access latency as low as 19usec, Exadata X8M is now up to 10x faster than before, boosting performance of a wide range of OLTP and Analytical applications. What are you going to do with all that speed? You can consolidate more workloads with less infrastructure and fewer CPUs, and ultimately lower your total cost of ownership. To boost Exadata’s consolidation capabilities, ExaCS X8M will have almost double the usable memory (1,390 GB per server) and double the local disk space for Oracle Homes.

But Exadata Cloud Service X8M is not just catching up to its sibling Exadata Cloud@Customer, it is redefining elasticity in the cloud. Exadata Cloud Service now supports independent scaling of compute and storage up to 32 compute servers and 64 storage cells. This provides three key benefits to customers:

◉ No workload is too large: Customers can scale to support huge workloads with up to 1,600 cores, 44 TB memory, 96 TB PMEM, 1.6 PB Flash, and 2.5 PB of usable data (25 PB after 10x HCC Compression).

◉ Right-sized systems: Customers can choose the optimum mix of storage and compute. They are no longer required to deploy compute and storage in fixed ratios, eliminating the need to pay for unnecessary infrastructure. They can choose what is best for their workload.

◉ On-demand Elasticity: Grow your system at any time by adding additional compute and storage servers. Pay for that capacity only when you need it. 

To make this work, we have redefined the concept of an Exadata in the cloud. Gone are the fixed rack sizes (Quarter, Half, and Full). Now, you start out with a minimal elastic configuration (happens to be the same as a quarter rack) of two database servers and three storage servers, and you add database and storage servers one or more at a time as required. The Oracle cloud takes care of provisioning these servers from pools of RoCE interconnected compute and storage, building the ideal system as you go.

DB Exam Prep, Database Tutorial and Material, Database Learning, Database Certification

Start with a small quarter rack and grow it over time as your business grows. If you need more storage, add more. If your system needs additional compute to remain balanced, add compute. Pay for this additional infrastructure only when you need it. With maximum configurations of 32 compute servers and 64 storage cells, be confident that the system can scale to meet future growth.

Exadata Cloud Service X8M will be available globally in the next few weeks. Pricing is unchanged from Exadata Cloud Service X8—you get persistent memory and the new faster RoCE interconnect for no additional charge. You can try it out with little investment as we allow you to provision a system for as little as 48 hours.

Related Posts

0 comments:

Post a Comment