As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
With the advent of the big data era, the explosive growth of data volume has put enormous computing and storage pressure on electric power companies. As a powerful technical means, supercomputing clouds are widely used in data processing, storage, and online services. However, critical services in supercomputing clouds are often deployed with overprovisioned resources to ensure the quality of service for users, resulting in significant energy consumption and additional costs. At the same time, insufficient resources for service provisioning can lead to performance degradation and service violations. To this end, this paper proposes an automatic resource scaling management system on electric power supercomputing clouds. Specifically, the proposed system is based on Transformer’s long-sequence prediction model to predict the future load intensity of the service and calculate the number of instances required by the service in the future through the runtime service requirement estimation component, thus automatically scaling resources and minimizing resource costs. Experimental results show that our system achieves the best scaling behavior based on performance metrics and the lowest cost overhead compared to strong baselines.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.