• Corpus ID: 231627464

CPU Scheduling in Data Centers Using Asynchronous Finite-Time Distributed Coordination Mechanisms

  title={CPU Scheduling in Data Centers Using Asynchronous Finite-Time Distributed Coordination Mechanisms},
  author={Andreas Grammenos and Themistoklis Charalambous and Evangelia Kalyvianaki},
We propose an asynchronous iterative scheme that allows a set of interconnected nodes to distributively reach an agreement within a pre-specified bound in a finite number of steps. While this scheme could be adopted in a wide variety of applications, we discuss it within the context of task scheduling for data centers. In this context, the algorithm is guaranteed to approximately converge to the optimal scheduling plan, given the available resources, in a finite number of steps. Furthermore, by… 
Optimal CPU Scheduling in Data Centers via a Finite-Time Distributed Quantized Coordination Mechanism
A fast distributed iterative algorithm, which operates over a large scale network of nodes, and allows each of the interconnected nodes to reach agreement to an optimal solution in a finite number of time steps is proposed.
Fast Quantized Average Consensus over Static and Dynamic Directed Graphs
This paper presents and analyzes a distributed averaging algorithm which operates exclusively with quantized values and extends the operation of the algorithm to achieve finitetime convergence in the presence of a dynamic directed communication topology subject to some connectivity conditions.
An Asynchronous Approximate Distributed Alternating Direction Method of Multipliers in Digraphs
This work proposes an algorithm, herein called Asynchronous Approximate Distributed Alternating Direction Method of Multipliers (AsyAD-ADMM), using finite-time asynchronous approximate ratio consensus, to solve the multi-node convex optimization problem, in which every node performs iterative computations and exchanges information with its neighbors asynchronously.
Distributed Finite-Time Privacy-Preserving Optimal Allocation of Test Kits for Controlling Pandemic Spreading
This paper proposes a distributed privacy-preserving optimal resource allocation algorithm with efficient (i.e., quantized) communication over a directed communication network and proves that, under specific conditions on the network topology, nodes are able to preserve the privacy of their initial state.


Average Consensus in the Presence of Delays in Directed Graph Topologies
A protocol is proposed that ensures asymptotic consensus to the exact average, despite the presence of arbitrary (but bounded) delays in the communication links, and its proof of correctness relies on the weak convergence of a backward product of column stochastic matrices.
Distributed Finite-Time Average Consensus in Digraphs in the Presence of Time Delays
This paper introduces a distributed protocol that allows nodes to find the exact average of the initial values in a finite and minimum number of steps on interconnection topologies described by strongly connected directed graphs (digraphs).
Quincy: fair scheduling for distributed computing clusters
It is argued that data-intensive computation benefits from a fine-grain resource sharing model that differs from the coarser semi-static resource allocations implemented by most existing cluster computing architectures.
Finite-Time Approximate Consensus and Its Application to Distributed Frequency Regulation in Islanded AC Microgrids
The approach to achieving finite-time approximate consensus builds upon ratio consensus, a distributed iterative algorithm in which each node maintains two state variables where the ratio of the states converges asymptotically to a constant that is equal for all the nodes.
Apollo: Scalable and Coordinated Scheduling for Cloud-Scale Computing
Efficiently scheduling data-parallel computation jobs over cloud-scale computing clusters is critical for job performance, system throughput, and resource utilization. It is becoming even more
Asynchronous Distributed ADMM for Large-Scale Optimization—Part I: Algorithm and Convergence Analysis
This paper proposes an asynchronous distributed ADMM (AD-ADMM), which can effectively improve the time efficiency of distributed optimization, and analyzes the convergence conditions of the AD- ADMM, under the popular partially asynchronous model, which is defined based on a maximum tolerable delay of the network.
High Throughput Data Center Topology Design
This paper presents the first non-trivial upper-bound on network throughput under uniform traffic patterns for any topology with identical switches, and shows that random graphs achieve throughput surprisingly close to this bound, within a few percent at the scale of a few thousand servers.
Firmament: Fast, Centralized Cluster Scheduling at Scale
Firmament is described, a centralized scheduler that scales to over ten thousand machines at sub-second placement latency even though it continuously reschedules all tasks via a min-cost max-flow (MCMF) optimization, and exceeds the placement quality of four widely-used centralized and distributed schedulers on a real-world cluster.
Distributed Stopping Criterion for Consensus in the Presence of Delays
An algorithm for average consensus is developed that utilizes a distributive stopping criterion, based on maximum and minimum consensus, where no centralized coordination is needed on how each agent weights its neighbor's values.
On the convergence of the max-consensus protocol with asynchronous updates
A theoretical setting is proposed to prove the convergence of the asynchronous max-consensus protocol and an upper bound on the convergence time of the max- Consensus protocol in asynchronous networks is provided.