Performance in the Cloud is Measured by Cost

DateJun 9, 2024

The video narrates how a customer enhanced performance and reduced costs on their cloud subscription by clearly defining their goals. This clarity allowed for precise sizing and configuration of their database and application. The customer transitioned their GL reporting batch on PeopleSoft Financials to Exadata Cloud-at-Customer, implementing various database features like resource manager plans, partitioning, materialized views, compression, and In-Memory, which significantly improved efficiency.

Initially, their on-premises hardware setup involved guessing future needs, leading to over-provisioning and excess costs. In contrast, the cloud environment allowed for a pay-as-you-go model, emphasizing the importance of efficient CPU usage to manage costs effectively.

A vital strategy involved setting performance goals, such as ensuring reports were ready for users in Continental Europe by the start of the working day, without disrupting the Asia-Pacific users. This led to prioritizing processes and optimizing resource allocation, which played a crucial role in reducing cloud costs.

By leveraging features like SQL profiles and disabling parallelism, the customer reduced CPU usage further. The implementation of the resource manager helped in managing workloads and ensuring that high-priority tasks were given the necessary resources, while lower-priority tasks were deferred.

Overall, the video illustrates how strategic planning and resource management can lead to significant cost savings and performance improvements in a cloud environment, highlighting the shift from capital expenditure to operational expenditure and the need for continuous optimization in cloud resource usage.

Leave a Reply