/pdf/a-break-in-the-clouds-towards-a-cloud-definition-svovsw1r5a.pdf

A Break in the Clouds: Towards a Cloud Definition

The Grid 2: Blueprint for a New Computing Infrastructure

Google's Borg system is a cluster manager that runs hundreds of thousands of jobs, from many thousands of different applications, across a number of clusters each with up to tens of thousands of machines. It achieves high utilization by combining admission control, efficient task-packing, over-commitment, and machine sharing with process-level performance isolation. It supports high-availability applications with runtime features that minimize fault-recovery time, and scheduling policies that reduce the probability of correlated failures. Borg simplifies life for its users by offering a declarative job specification language, name service integration, real-time job monitoring, and tools to analyze and simulate system behavior. We present a summary of the Borg system architecture and features, important design decisions, a quantitative analysis of some of its policy decisions, and a qualitative examination of lessons learned from a decade of operational experience with it.

/pdf/large-scale-cluster-management-at-google-with-borg-4i7d9qnw5x.pdf

Large-scale cluster management at Google with Borg

To better understand the challenges in developing effective cloud-based resource schedulers, we analyze the first publicly available trace data from a sizable multi-purpose cluster. The most notable workload characteristic is heterogeneity: in resource types (e.g., cores:RAM per machine) and their usage (e.g., duration and resources needed). Such heterogeneity reduces the effectiveness of traditional slot- and core-based scheduling. Furthermore, some tasks are constrained as to the kind of machine types they can use, increasing the complexity of resource assignment and complicating task migration. The workload is also highly dynamic, varying over time and most workload features, and is driven by many short jobs that demand quick scheduling decisions. While few simplifying assumptions apply, we find that many longer-running jobs have relatively stable resource utilizations, which can help adaptive resource schedulers.

http://www.istc-cc.cmu.edu/publications/papers/2012/googletrace-socc2012.pdf

Heterogeneity and dynamicity of clouds at scale: Google trace analysis

Network-based cloud computing is rapidly expanding as an alternative to conventional office-based computing. As cloud computing becomes more widespread, the energy consumption of the network and computing resources that underpin the cloud will grow. This is happening at a time when there is increasing attention being paid to the need to manage energy consumption across the entire information and communications technology (ICT) sector. While data center energy use has received much attention recently, there has been less attention paid to the energy consumption of the transmission and switching networks that are key to connecting users to the cloud. In this paper, we present an analysis of energy consumption in cloud computing. The analysis considers both public and private clouds, and includes energy consumption in switching and transmission as well as data processing and data storage. We show that energy consumption in transport and switching can be a significant percentage of total energy consumption in cloud computing. Cloud computing can enable more energy-efficient use of computing power, especially when the computing tasks are of low intensity or infrequent. However, under some circum- stances cloud computing can consume more energy than conventional computing where each user performs all com- puting on their own personal computer (PC).

/pdf/green-cloud-computing-balancing-energy-in-processing-storage-2z2x5rjq9b.pdf

Green Cloud Computing: Balancing Energy in Processing, Storage, and Transport For processing large amounts of data, management and switching of communications may contribute significantly to energy consumption and cloud computing seems to be an alternative to office-based computing.

By leveraging virtual machine (VM) technology, we optimize cloud system performance based on refined resource allocation, in processing user requests with composite services. Our contribution is three-fold. (1) We devise a VM resource allocation scheme with a minimized processing overhead for task execution. (2) We comprehensively investigate the best-suited task scheduling policy with different design parameters. (3) We also explore the best-suited resource sharing scheme with adjusted divisible resource fractions on running tasks in terms of Proportional-share model (PSM), which can be split into absolute mode (called AAPSM) and relative mode (RAPSM). We implement a prototype system over a cluster environment deployed with 56 real VM instances, and summarized valuable experience from our evaluation. As the system runs in short supply, lightest workload first (LWF) is mostly recommended because it can minimize the overall response extension ratio (RER) for both sequential-mode tasks and parallel-mode tasks. In a competitive situation with over-commitment of resources, the best one is combining LWF with both AAPSM and RAPSM. It outperforms other solutions in the competitive situation, by 16 $\;+\;$ % w.r.t. the worst-case response time and by 7.4 $\;+\;$ % w.r.t. the fairness.

Optimization of Composite Cloud Service Processing with Virtual Machines

In large-scale Grid computing environments, providing fault-tolerance is required for both scientific computation and file-sharing to increase their reliability. In previous works, several mechanisms were proposed for the Grids or distributed computing systems. However, some of them used only space redundancy (hardware replication), and others used only time redundancy (checkpointing and rollback). For this reason, the existing mechanisms are inefficient in terms of their resource utilization on the Grids. The main goal of ART is reducing the number of replications by using checkpointing and rollback scheme for each replication. In ART, the minimum number of replications is adaptively selected based on analysis of probability of successful execution within the given deadline and reliability requirement of each task. Our simulation results show that ART can significantly reduce the number of replications and improve scalability compared with existing mechanisms.

/pdf/using-replication-and-checkpointing-for-reliable-task-4r5z9c3ai0.pdf

Using replication and checkpointing for reliable task management in computational Grids

In the age of cloud, Grid, P2P, and volunteer distributed computing, large-scale systems with tens of thousands of unreliable hosts are increasingly common. Invariably, these systems are composed of heterogeneous hosts whose individual availability often exhibit different statistical properties (for example stationary versus non-stationary behaviour) and fit different models (for example Exponential, Weibull, or Pareto probability distributions). In this paper, we describe an effective method for discovering subsets of hosts whose availability have similar statistical properties and can be modelled with similar probability distributions. We apply this method with about 230,000 host availability traces obtained from a real large-scale Internet-distributed system, namely SETI@home. We find that about 34% of hosts exhibit availability that is a truly random process, and that these hosts can often be modelled accurately with a few distinct distributions from different families. We believe that this characterization is fundamental in the design of stochastic scheduling algorithms across large-scale systems where host availability is uncertain.

https://hal.inria.fr/inria-00375624/document

Mining for Availability Models in Large-Scale Distributed Systems:A Case Study of SETI@home

Desktop grids, which use the idle cycles of many desktop PC's, are currently the largest distributed systems in the world. Despite the popularity and success of many desktop grid projects, the heterogeneity and volatility of hosts within desktop grids has been poorly understood. Yet, host characterization is essential for accurate simulation and modelling of such platforms. In this paper, we present application-level traces of four real desktop grids that can be used for simulation and modelling purposes. In addition, we describe aggregate and per host statistics that reflect the heterogeneity and volatility of desktop grid resources.

https://hal.inria.fr/inria-00000994/document

Resource Availability in Enterprise Desktop Grids

Since the early 1990's, the largest distributed computing systems in the world have been desktop grids, which use the idle cycles of mainly desktop PC's to support large scale computation. Despite the enormous computing power offered by such systems, the range of supportable applications is largely limited to task parallel, compute-bound, and high-throughput applications. This limitation is mainly because of the heterogeneity and volatility of the underlying resources, which are shared with the desktop users. Our work focuses on broadening the applications supportable by desktop grids, and in particular, we focus on the development of scheduling heuristics to enable rapid turnaround for short-lived applications. 
To that end, the contributions of this dissertation are as follows. First, we measure and characterize four real enterprise desktop grid systems; such characterization is essential for accurate modelling and simulation. Second, using the characterization, we design scheduling heuristics that enable rapid application turnaround. These heuristics are based on three scheduling techniques, namely resource prioritization, resource exclusion, and task replication. We find that our best heuristic uses relatively static resource information for prioritization and exclusion, and reactive task replication to achieve performance within a factor of 1.7 of optimal. Third, we implement our best heuristic in a real desktop grid system to demonstrate its feasibility.

Derrick Kondo

Papers

Optimization of Composite Cloud Service Processing with Virtual Machines

Using replication and checkpointing for reliable task management in computational Grids

Mining for Availability Models in Large-Scale Distributed Systems:A Case Study of SETI@home

Resource Availability in Enterprise Desktop Grids

Scheduling task parallel applications for rapid turnaround on desktop grids