Optimal CPU Scheduling in Data Centers via a Finite-Time Distributed Quantized Coordination Mechanism

ABSTRACTTask scheduling algorithms for data centers must be capable of making instantaneous decisions based on the current state of the system. However, due to information limitations, these scheduling algorithms often fail to achieve optimal scheduling plans. To address the information bottleneck faced in DAG (Directed Acyclic Graph) task scheduling within data centers, this paper proposes a deep reinforcement learning scheduling model based on a DAG attention mechanism. This model utilizes the attention mechanism to capture the potential relationships between dependent tasks, thereby improving scheduling efficiency and system performance under limited information conditions. The experimental results indicate that our DAG attention mechanism can significantly reduce makespan.

Existing flow scheduling schemes in Data Center Network (DCN) are designed mainly to minimize the flow complete time (FCT) of short flows and do not consider optimizing the FCT of latency-sensitive long flows (e.g. VR video streaming, interactive artificial intelligence questiona and the information-agnostic scheme (i.e. PIAS), which is based on the premise that flow size is not known a priori, requires a central server, causing a poor scalability in large network scales. Given the limitations of existing solutions, in this paper, we propose a distributed information-agnostic flow scheduling scheme (DIAS), which minimizes the FCT of both short flows and latency-sensitive long flows. In DIAS, packets are forwarded complying with their priorities, which are determined based on packets wait-time that is defined as staying time in end hosts' send buffers, and the longer a packet stays in a send buffer, the lower its priority. Meanwhile, instead of utilizing a central server to collect traffic load information, each switch feeds traffic load information which is used to adjust the thresholds of determining packets priority back to end hosts via ACK packets. The experimental results in ns-3 simulator show that DIAS reduces FCT by up to 54.7% and 50.1 % over DCTCP and $L^{2}\mathbf{DCT}$ , respectively. Besides, DIAS ensures a smaller FCT of latency-sensitive long flows and performs better than PIAS.

Hybrid ant genetic algorithm for efficient task scheduling in cloud data centers

Spatio-temporal thermal-aware job scheduling to minimize energy consumption in virtualized heterogeneous data centers

We consider a class of k-server optimal task scheduling problems partitioning and scheduling N tasks with various real-time constrains and work loads on k servers with convex task processing cost function so as to minimize the total task processing cost while still guaranteeing satisfaction of all time constraints. This class has broad expressing power for practical scheduling problems in several areas such as real-time multimedia wireless transmission , CPU energy conservation, and warehouse order processing management, et. al. Our formulation is quite general such that most previous works can be readily reduced to a special case of the presented k-server optimal task scheduling problem. We show that, when k = 1, optimal solution can be obtained in computational complexity of O(N) and the corresponding optimal scheduling problem is equivalent to finding the shortest 2D Euclidean distance between two vertices inside a well-defined 2D polygon. However, when k 2, the optimal scheduling problem can be demonstrated to be NP-hard by reducing it to a well-known NP-complete bin-packing problem. Therefore, we conclude no polynomial time algorithm exists for a general k-server optimal task scheduling problem. We then construct approximation algorithms to solve the presented k-server problem in a practical way and illustrate its performance by simulation results and analysis.

This work studies the power optimal dynamic scheduling problem in multi-channel multi-user wireless access networks. Users have quality-of-service (QoS) requirements on the minimum rates with statistical delay guarantees. Only one user is allowed to transmit over a channel in a given time slot. This work considers two scenarios: homogeneous and heterogeneous users. For the former scenario, the optimal scheduling policy can be derived, and an online scheduling algorithm for the optimal policy is proposed using online time-averaging without requiring a-priori known fading statistics. For the latter scenario, the optimal scheduling problem is combinatorially hard; hence, even when the fading statistics are available, computing the optimal policy is intractable. Consequently, this work develops a sub-optimal online scheduling algorithm with linear complexity which does not require a-priori known fading statistics. Moreover, the scheduling algorithm satisfies the QoS constraints for the users. Illustrative results demonstrate the performance of the proposed scheduling algorithms in various settings.

Distributed Average Consensus under Quantized Communication via Event-Triggered Mass Splitting

As the demand for large scale computing is rapidly increasing to serve billions of users across the world, more powerful and densely packed server configurations are being used. Often in developing countries, and in small and medium enterprises, it is hard to place such servers in sound-proof server rooms. Hence, servers are typically placed in close proximity to employees. The noise from the cooling fans in servers adversely affects employees' health, and reduces their productivity. In this paper, we provide a framework for computer architects to measure the acoustic profile in a data center along with the temperature profile, and estimate the sound power levels at points of interest. Additionally, we studied the noise levels obtained upon using algorithms targeted at homogenizing the temperature profile. We found that these algorithms result in high noise levels, sometimes above the permissible levels. So, we propose two heuristics to redistribute workloads in a data center such that noise can be reduced at certain target locations. We obtain a noise reduction of 2-13 dB when compared with uniform workload distribution, and upto 16 dB as compared to temperature aware workload placement, with a reduction of at least 5-6 dB in 75% of the cases. The performance overhead is limited to 1%.

Communication in data centers often involves many parallel flows that all share the same performance goal. A useful abstraction, coflow, is proposed to express the communication requirements of prevalent data parallel paradigms. The multiple coflow routing and scheduling problem faces challenges when deriving a good theoretical performance ratio because coexisting coflows will compete for the same network resources such as link bandwidths. In this paper, we focus on the coflow problem in the most popular data center infrastructure: the Leaf-Spine topology. We first formulate the problem and study the path selection issue on this two-tier structure. In order to minimize the average coflow completion time (CCT), we propose a Multi- hop Coflow Routing and Scheduling strategy (MCRS) for inter-coflows and intra-coflows and prove that our method has a reasonably good competitive ratio. Extensive experiments show that MCRS outperforms other state-of- art schemes.

The problem of task scheduling with communication delays (\(P|prec,c_{ij}|C_{\mathrm {max}}\)) is NP-hard, and therefore solutions are often found using a heuristic method. However, an optimal schedule can be very useful in applications such as time critical systems, or as a baseline for the evaluation of heuristics. Branch-and-bound algorithms such as A* have previously been shown to be a promising approach to the optimal solving of this problem, using a state-space model which we refer to as exhaustive list scheduling. However, this model suffers from the possibility of producing high numbers of duplicate states. In this paper we define a new state-space model in which we divide the problem into two distinct subproblems: first we decide the allocations of all tasks to processors, and then we order the tasks on their allocated processors to produce a complete schedule. This two-phase state-space model offers no potential for the production of duplicates. An empirical evaluation shows that the use of this new state-space model leads to a marked reduction in the number of states considered by an A* search in many cases, particularly for task graphs with a high communication-to-computation ratio. With additional refinement, and the development of specialised pruning techniques, the performance of this state-space model could be improved.

Eventually-consistent federated scheduling for data center workloads

This paper explores integrating biomechanics data and bio-inspired models to enhance efficiency in business administration, focusing on task scheduling, resource allocation, and workflow optimization. Biomechanics, traditionally applied in fields such as healthcare and sports, is used to analyze human movement and physical strain in business processes, particularly in physically demanding environments like manufacturing and logistics. Bio-inspired models, such as Genetic Algorithms (GA) and Particle Swarm Optimization (PSO), are applied to solve complex optimization problems in resource management and task scheduling. The study presents three case studies to demonstrate the practical application of these methodologies: (1) workflow optimization in a manufacturing environment using biomechanics data to reduce physical strain and improve task completion times; (2) resource allocation in Supply Chain Management (SCM) using PSO to minimize transportation and labor costs while improving warehouse utilization and delivery times; and (3) task scheduling in an office environment using GA to enhance task efficiency, workload distribution, and employee satisfaction. The case study results demonstrate the practical application of these methodologies: (1) a 21.6% reduction in shoulder joint strain and an 18.2% improvement in task completion time in a manufacturing setting; (2) a 16.1% reduction in transportation costs and an 18.6% improvement in warehouse utilization in SCM using PSO; and (3) a 17.6% decrease in makespan and a 29.8% improvement in workload distribution through GA-based task scheduling in an office environment. These findings underscore the potential of combining human-centered biomechanics data with bio-inspired optimization models to improve operational efficiency, employee well-being, and cost-effectiveness significantly.

A metaheuristic method for joint task scheduling and virtual machine placement in cloud data centers

The proposed work concentrates on the networking facets of sensor-cloud infrastructures—one of the first attempts of its kind. In a sensor-cloud, multiple sets of physical sensor nodes that are activated based on an application demand, in turn give rise to multiple distinct virtual sensors (VSs). The VSs are considered to span across multiple geographical regions; thereby, depositing the data (from each of the VS) to the closest cloud data center (DC). Quite obviously, multiple geospatial DCs get involve with an application. However, the principle of sensor-cloud is to store and conglomerate the data from various VSs, before they can be provisioned as Sensors-as-a-Service (Se-aaS). The assortation of data occurs within a single Virtual Machine (VM) (or in some cases multiple VMs) residing inside a particular DC. This work addresses the problem of scheduling a particular DC that congregates data from various VSs, and transmit the same to the end-user application. The work follows the general pairwise choice framework of the Optimal Decision Rule. The scheduling of the DC is performed under several network constraints, such as data migration cost, data delivery cost, and service delay of an application that ensures the preservation of the Quality-of-Service (QoS) and maintenance of the user satisfaction. The work quantifies the effective QoS of Se-aaS and determines an optimal decision rule for electing a particular DC. While arriving at a collective decision, the work incorporates the fallible decision making ability of a DC; thereby, excluding the loss of generality. Experimental results depict that the proposed algorithm for generating the optimal decision rule finds applicability in real-time cloud computing scenarios.

In distributed computing frameworks like MapReduce, Spark, and Dyrad, a coflow is a set of flows transferring data between two stages of a job. The job cannot start its next stage unless all flows in the coflow finish. To improve the execution performance of such a job, it is crucial to reduce the completion time of a coflow, as it can contribute more than 50 percent of the job completion time. While several coflow schedulers have been proposed, we observe that routing, as a factor greatly impacting the Coflow Completion Time (CCT), has not been well considered. In this article, we focus on the coflow scheduling problem and jointly consider routing and bandwidth allocation. We begin by providing an analytical solution to the problem of optimal bandwidth allocation with pre-determined routes. In the following, we formulate the problem of scheduling a single coflow as a Non-linear Mixed Integer Programming problem and present its relaxed convex optimization problem. We further propose two algorithms, CoRBA and its simplified version: CoRBA-fast that solve the single coflow scheduling problem with a joint consideration of routing and bandwidth allocation. Lastly, to address multiple coflows in online scheduling, we propose an online scheduler named OnCoRBA. By comparing with the start-of-the-art algorithms and schedulers via simulations, we demonstrate that CoRBA and CoRBA-fast reduce the CCT by 30-400 percent and the OnCoRBA scheduler reduces the average online CCT by 20-230 percent. In addition, CoRBA-fast can be hundreds times faster than CoRBA with around 8 percent performance degradation compared to CoRBA, which makes the use of CoRBA-fast very appropriate in practice.