English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Problem Definition: Allocating sufficient capacity to cloud services is achallenging task, especially when demand is time-varying, heterogeneous,contains batches, and requires multiple types of resources for processing. Inthis setting, providers decide whether to reserve portions of their capacity toindividual job classes or to offer it in a flexible manner.Methodology/results: In collaboration with Huawei Cloud, a worldwide providerof cloud services, we propose a heuristic policy that allocates multiple typesof resources to jobs and also satisfies their pre-specified service levelagreements (SLAs). We model the system as a multi-class queueing network withparallel processing and multiple types of resources, where arrivals (i.e.,virtual machines and containers) follow time-varying patterns and require atleast one unit of each resource for processing. While virtual machines leave ifthey are not served immediately, containers can join a queue. We introduce adiffusion approximation of the offered load of such system and investigate itsfidelity as compared to the observed data. Then, we develop a heuristicapproach that leverages this approximation to determine capacity levels thatsatisfy probabilistic SLAs in the system with fully flexible servers.Managerial Implications: Using a data set of cloud computing requests over arepresentative 8-day period from Huawei Cloud, we show that our heuristicpolicy results in a 20% capacity reduction and better service quality ascompared to a benchmark that reserves resources. In addition, we show that thesystem utilization induced by our policy is superior to the benchmark, i.e., itimplies less idling of resources in most instances. Thus, our approach enablescloud operators to both reduce costs and achieve better performance.

Capacity Allocation for Clouds with Parallel Processing, Batch Arrivals, and Heterogeneous Service Requirements