Search

Edge-cloud applications are rapidly prevailing in recent years and pose the challenge of using both resource-strenuous edge devices and elastic cloud resources under dynamic workloads. Efficient resource allocation on edge-cloud jobs via cluster schedulers (e.g. Kubernetes/Volcano scheduler) is essential to guarantee their performance, e.g. tail latency, and such allocation is sensitive to scheduler configurations such as applied scheduling algorithms and task restart/discard policy. Deep reinforcement learning (DRL) is increasingly applied to optimize scheduling decisions. However, DRL faces the conundrum of achieving high rewards at a dauntingly long training time (e.g. hours or days), making it difficult to tune the scheduler configurations online in accordance to dynamically changing edge-cloud workloads and resources. For such an issue, this paper proposes EdgeTuner, a fast scheduler configuration tuning approach that efficiently leverages DRL to reduce tail latency of edge-cloud jobs. The enabling feature of EdgeTuner is to effectively simulate the execution of edge-cloud jobs under different scheduler configurations and thus quickly estimate these configurations’ influence on job performance. The simulation results allow EdgeTuner to timely train a DRL agent in order to properly tune scheduler configurations in dynamic edge-cloud environment. We implement EdgeTuner in both Kubernetes and Volcano schedulers and extensively evaluate it on real workloads driven by Alibaba production traces. Our results show that EdgeTuner outperforms prevailing scheduling algorithms by achieving much lower tail latency while accelerating DRL training speed by an average of 151.63x., Data-Intensive Systems

Enhanced nitrogen (N) and phosphorus (P) fertilizer applications have strongly elevated agricultural productivity, but also caused significant soil acidification (by N) and increased N and P losses to water bodies leading to eutrophication. There is an urgent need to integrate soil acidification amendment and nutrient management strategies to stimulate sustainable soil use and minimize environmental risks. This study aimed to derive the optimal combination of fertilizer, manure and lime to mitigate soil acidification and minimize nutrient losses by assessing the impacts of different nutrient management strategies on soil pH and nutrient budgets at a field level. The dynamic soil acidification model VSD+ was calibrated on measured pH changes in a long-term fertilization experiment (1990–2018), in which variable amounts of inorganic N, P and potassium (K), lime (L) and organic manure (M) were applied annually to a maize-wheat rotation on red soil. Adding NPK inorganic fertilizers only sharply decreased the soil pH by 1.5 units within 10 years. The pH remained relatively constant, however, in the experiment in which 70% of the N given by fertilizer was replaced by manure (NPKM) while lime addition after 20 years of NPK addition (NPKL) increased the pH by approximately 1.5 units. Historic soil pH changes were well reconstructed by VSD+. The model quantified that nitrogen transformations contributed up to 89% of the total acidity production in the NPK treatment, which could be fully neutralized by replacing 70% of the total N fertilizer input with manure. Scenarios were assessed to explore the future impacts of adapted nutrient management on soil acidification and nutrient surplus. Balancing the N and P input with crop demand (BNP) significantly reduces the N and P surplus but hardly affects the acidification rate compared to Business as Usual (BAU, equal to the NPK treatment). Balancing the N input and substituting N fertilizer with manure (BM) counteract acidification

Edge-cloud jobs are rapidly prevailing in many application domains, posing the challenge of using both resource-strenuous edge devices and elastic cloud resources. Efficient resource allocation on such jobs via scheduling algorithms is essential to guarantee their performance, e.g. latency. Deep reinforcement learning (DRL) is increasingly adopted to make scheduling decisions but faces the conundrum of achieving high rewards at a low training overhead. It is unknown if such a DRL can be applied to timely tune the scheduling algorithms that are adopted in response to fast changing workloads and resources. In this paper, we propose EdgeTuner to effectively leverage DRL to select scheduling algorithms online for edge-cloud jobs. The enabling features of EdgeTuner are sophisticated DRL model that captures complex dynamics of Edge-Cloud jobs/tasks and an effective simulator to emulate the response times of short-running jobs in accordance to dynamically changing scheduling algorithms. EdgeTuner trains DRL agents offline by directly interacting with the simulator. We implement EdgeTuner on Kubernetes scheduler and extensively evaluate it on Kubernetes cluster testbed driven by the production traces. Our results show that EdgeTuner outperforms prevailing scheduling algorithms by achieving significant lower job response time while accelerating DRL training speed by more than 180x., Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Data-Intensive Systems