arXiv subjects

Dimitrios S. Nikolopoulos

Publications and source records attributed to Dimitrios S. Nikolopoulos.

At least 37 records · Page 2Linked to original sources

arXiv2020

DYVERSE: DYnamic VERtical Scaling in Multi-tenant Edge Environments

Multi-tenancy in resource-constrained environments is a key challenge in Edge computing. In this paper, we develop 'DYVERSE: DYnamic VERtical Scaling in Edge' environments, which is the first light-weight and dynamic vertical scaling mechanism for managing resources allocated to applications for facilitating multi-tenancy in Edge environments. To enable dynamic vertical scaling, one static and three dynamic priority management approaches that are workload-aware, community-aware and system-aware, respectively are proposed. This research advocates that dynamic vertical scaling and priority management approaches reduce Service Level Objective (SLO) violation rates. An online-game and a face detection workload in a Cloud-Edge test-bed are used to validate the research. The merits of DYVERSE is that there is only a sub-second overhead per Edge server when 32 Edge servers are deployed on a single Edge node. When compared to executing applications on the Edge servers without dynamic vertical scaling, static priorities and dynamic priorities reduce SLO violation rates of requests by up to 4% and 12% for the online game, respectively, and in both cases 6% for the face detection workload. Moreover, for both workloads, the system-aware dynamic vertical scaling method effectively reduces the latency of non-violated requests, when compared to other methods.

cs.DC↗

arXiv2019

Implementing Efficient Message Logging Protocols as MPI Application Extensions

Message logging protocols are enablers of local rollback, a more efficient alternative to global rollback, for fault tolerant MPI applications. Until now, message logging MPI implementations have incurred the overheads of a redesign and redeployment of an MPI library, as well as continued performance penalties across various kernels. Successful research efforts for message logging implementations do exist, but not a single one of them can be easily deployed today by more than a few experts. In contrast, in this work we build efficient message logging capabilities on top of an MPI library with no message logging capabilities; we do so for two different HPC kernels, one with a global exchange pattern (CG), and one with a neighbourhood exchange pattern (LULESH). While our library of choice ULFM detects failure and recovers MPI communicators, we build on that to then restore the intra- and inter-process data consistency of both applications. This task turns out to be challenging, and we present the methodology for doing so in this work. In the end, we achieve message logging capabilities for each kernel, without the need for an actual message logging runtime underneath. On the performance side, we match state-of-the-art solutions and (a) eliminate event logging and the event logger component altogether, and (b) design a hybrid protocol, which gracefully shifts between global and local rollback, depending on the available payload logging memory. Such a hybrid protocol between local and global rollback has not been previously proposed to our knowledge. Our extensions span a few hundred lines of code for each kernel, are open-sourced, and enable local and global rollback after process failure.

cs.DC↗

arXiv2018

RADS: Real-time Anomaly Detection System for Cloud Data Centres

Cybersecurity attacks in Cloud data centres are increasing alongside the growth of the Cloud services market. Existing research proposes a number of anomaly detection systems for detecting such attacks. However, these systems encounter a number of challenges, specifically due to the unknown behaviour of the attacks and the occurrence of genuine Cloud workload spikes, which must be distinguished from attacks. In this paper, we discuss these challenges and investigate the issues with the existing Cloud anomaly detection approaches. Then, we propose a Real-time Anomaly Detection System (RADS) for Cloud data centres, which uses a one class classification algorithm and a window-based time series analysis to address the challenges. Specifically, RADS can detect VM-level anomalies occurring due to DDoS and cryptomining attacks. We evaluate the performance of RADS by running lab-based experiments and by using real-world Cloud workload traces. Evaluation results demonstrate that RADS can achieve 90-95% accuracy with a low false positive rate of 0-3%. The results further reveal that RADS experiences fewer false positives when using its window-based time series analysis in comparison to using state-of-the-art average or entropy based analysis.

cs.DC↗

arXiv2018

VEBO: A Vertex- and Edge-Balanced Ordering Heuristic to Load Balance Parallel Graph Processing

Graph partitioning drives graph processing in distributed, disk-based and NUMA-aware systems. A commonly used partitioning goal is to balance the number of edges per partition in conjunction with minimizing the edge or vertex cut. While this type of partitioning is computationally expensive, we observe that such topology-driven partitioning nonetheless results in computational load imbalance. We propose Vertex- and Edge-Balanced Ordering (VEBO): balance the number of edges and the number of unique destinations of those edges. VEBO optimally balances edges and vertices for graphs with a power-law degree distribution. Experimental evaluation on three shared-memory graph processing systems (Ligra, Polymer and GraphGrind) shows that VEBO achieves excellent load balance and improves performance by 1.09x over Ligra, 1.41x over Polymer and 1.65x over GraphGrind, compared to their respective partitioning algorithms, averaged across 8 algorithms and 7 graphs.

cs.DC↗

arXiv2017

Intra-node Memory Safe GPU Co-Scheduling

GPUs in High-Performance Computing systems remain under-utilised due to the unavailability of schedulers that can safely schedule multiple applications to share the same GPU. The research reported in this paper is motivated to improve the utilisation of GPUs by proposing a framework, we refer to as schedGPU, to facilitate intra-node GPU co-scheduling such that a GPU can be safely shared among multiple applications by taking memory constraints into account. Two approaches, namely a client-server and a shared memory approach are explored. However, the shared memory approach is more suitable due to lower overheads when compared to the former approach. Four policies are proposed in schedGPU to handle applications that are waiting to access the GPU, two of which account for priorities. The feasibility of schedGPU is validated on three real-world applications. The key observation is that a performance gain is achieved. For single applications, a gain of over 10 times, as measured by GPU utilisation and GPU memory utilisation, is obtained. For workloads comprising multiple applications, a speed-up of up to 5x in the total execution time is noted. Moreover, the average GPU utilisation and average GPU memory utilisation is increased by 5 and 12 times, respectively.

cs.DC↗

arXiv2017

Edge-as-a-Service: Towards Distributed Cloud Architectures

We present an Edge-as-a-Service (EaaS) platform for realising distributed cloud architectures and integrating the edge of the network in the computing ecosystem. The EaaS platform is underpinned by (i) a lightweight discovery protocol that identifies edge nodes and make them publicly accessible in a computing environment, and (ii) a scalable resource provisioning mechanism for offloading workloads from the cloud on to the edge for servicing multiple user requests. We validate the feasibility of EaaS on an online game use-case to highlight the improvement in the QoS of the application hosted on our cloud-edge platform. On this platform we demonstrate (i) low overheads of less than 6%, (ii) reduced data traffic to the cloud by up to 95% and (iii) minimised application latency between 40%-60%.

cs.DC↗

arXiv2017

Power Modelling for Heterogeneous Cloud-Edge Data Centers

Existing power modelling research focuses not on the method used for developing models but rather on the model itself. This paper aims to develop a method for deploying power models on emerging processors that will be used, for example, in cloud-edge data centers. Our research first develops a hardware counter selection method that appropriately selects counters most correlated to power on ARM and Intel processors. Then, we propose a two stage power model that works across multiple architectures. The key results are: (i) the automated hardware performance counter selection method achieves comparable selection to the manual selection methods reported in literature, and (ii) the two stage power model can predict dynamic power more accurately on both ARM and Intel processors when compared to classic power models.

cs.PF↗

arXiv2017

ENORM: A Framework For Edge NOde Resource Management

Current computing techniques using the cloud as a centralised server will become untenable as billions of devices get connected to the Internet. This raises the need for fog computing, which leverages computing at the edge of the network on nodes, such as routers, base stations and switches, along with the cloud. However, to realise fog computing the challenge of managing edge nodes will need to be addressed. This paper is motivated to address the resource management challenge. We develop the first framework to manage edge nodes, namely the Edge NOde Resource Management (ENORM) framework. Mechanisms for provisioning and auto-scaling edge node resources are proposed. The feasibility of the framework is demonstrated on a PokeMon Go-like online game use-case. The benefits of using ENORM are observed by reduced application latency between 20% - 80% and reduced data transfer and communication frequency between the edge node and the cloud by up to 95\%. These results highlight the potential of fog computing for improving the quality of service and experience.

cs.DC↗

arXiv2017

Dependency-Aware Rollback and Checkpoint-Restart for Distributed Task-Based Runtimes

With the increase in compute nodes in large compute platforms, a proportional increase in node failures will follow. Many application-based checkpoint/restart (C/R) techniques have been proposed for MPI applications to target the reduced mean time between failures. However, rollback as part of the recovery remains a dominant cost even in highly optimised MPI applications employing C/R techniques. Continuing execution past a checkpoint (that is, reducing rollback) is possible in message-passing runtimes, but extremely complex to design and implement. Our work focuses on task-based runtimes, where task dependencies are explicit and message passing is implicit. We see an opportunity for reducing rollback for such runtimes: we explore task dependencies in the rollback, which we call dependency-aware rollback. We also design a new C/R technique, which is influenced by recursive decomposition of tasks, and combine it with dependency-aware rollback. We expect the dependency-aware rollback to cancel and recompute less tasks in the presence of node failures. We describe, implement and validate the proposed protocol in a simulator, which confirms these expectations. In addition, we consistently observe faster overall execution time for dependency-aware rollback in the presence of faults, despite the fact that reduced task cancellation does not guarantee reduced overall execution time.

cs.DC↗

arXiv2017

Feasibility of Fog Computing

As billions of devices get connected to the Internet, it will not be sustainable to use the cloud as a centralised server. The way forward is to decentralise computations away from the cloud towards the edge of the network closer to the user. This reduces the latency of communication between a user device and the cloud, and is the premise of 'fog computing' defined in this paper. The aim of this paper is to highlight the feasibility and the benefits in improving the Quality-of-Service and Experience by using fog computing. For an online game use-case, we found that the average response time for a user is improved by 20% when using the edge of the network in comparison to using a cloud-only model. It was also observed that the volume of traffic between the edge and the cloud server is reduced by over 90% for the use-case. The preliminary results highlight the potential of fog computing in achieving a sustainable computing model and highlights the benefits of integrating the edge of the network into the computing ecosystem.

cs.DC↗

arXiv2016

ALEA: Fine-grain Energy Profiling with Basic Block Sampling

Energy efficiency is an essential requirement for all contemporary computing systems. We thus need tools to measure the energy consumption of computing systems and to understand how workloads affect it. Significant recent research effort has targeted direct power measurements on production computing systems using on-board sensors or external instruments. These direct methods have in turn guided studies of software techniques to reduce energy consumption via workload allocation and scaling. Unfortunately, direct energy measurements are hampered by the low power sampling frequency of power sensors. The coarse granularity of power sensing limits our understanding of how power is allocated in systems and our ability to optimize energy efficiency via workload allocation. We present ALEA, a tool to measure power and energy consumption at the granularity of basic blocks, using a probabilistic approach. ALEA provides fine-grained energy profiling via statistical sampling, which overcomes the limitations of power sensing instruments. Compared to state-of-the-art energy measurement tools, ALEA provides finer granularity without sacrificing accuracy. ALEA achieves low overhead energy measurements with mean error rates between 1.4% and 3.5% in 14 sequential and parallel benchmarks tested on both Intel and ARM platforms. The sampling method caps execution time overhead at approximately 1%. ALEA is thus suitable for online energy monitoring and optimization. Finally, ALEA is a user-space tool with a portable, machine-independent sampling method. We demonstrate two use cases of ALEA, where we reduce the energy consumption of a k-means computational kernel by 37% and an ocean modelling code by 33%, compared to high-performance execution baselines, by varying the power optimization strategy between basic blocks.

cs.PF↗

arXiv2016

Challenges and Opportunities in Edge Computing

Many cloud-based applications employ a data centre as a central server to process data that is generated by edge devices, such as smartphones, tablets and wearables. This model places ever increasing demands on communication and computational infrastructure with inevitable adverse effect on Quality-of-Service and Experience. The concept of Edge Computing is predicated on moving some of this computational load towards the edge of the network to harness computational capabilities that are currently untapped in edge nodes, such as base stations, routers and switches. This position paper considers the challenges and opportunities that arise out of this new direction in the computing landscape.

cs.DC↗

arXiv2016

Myrmics: Scalable, Dependency-aware Task Scheduling on Heterogeneous Manycores

Task-based programming models have become very popular, as they offer an attractive solution to parallelize serial application code with task and data annotations. They usually depend on a runtime system that schedules the tasks to multiple cores in parallel while resolving any data hazards. However, existing runtime system implementations are not ready to scale well on emerging manycore processors, as they often rely on centralized structures and/or locks on shared structures in a cache-coherent memory. We propose design choices, policies and mechanisms to enhance runtime system scalability for single-chip processors with hundreds of cores. Based on these concepts, we create and evaluate Myrmics, a runtime system for a dependency-aware, task-based programming model on a heterogeneous hardware prototype platform that emulates a single-chip processor of 8 latency-optimized and 512 throughput-optimized CPUs. We find that Myrmics scales successfully to hundreds of cores. Compared to MPI versions of the same benchmarks with hand-tuned message passing, Myrmics achieves similar scalability with a 10-30% performance overhead, but with less programming effort. We analyze the scalability of the runtime system in detail and identify the key factors that contribute to it.

cs.DC↗

arXiv2016

BDDT-SCC: A Task-parallel Runtime for Non Cache-Coherent Multicores

This paper presents BDDT-SCC, a task-parallel runtime system for non cache-coherent multicore processors, implemented for the Intel Single-Chip Cloud Computer. The BDDT-SCC runtime includes a dynamic dependence analysis and automatic synchronization, and executes OpenMP-Ss tasks on a non cache-coherent architecture. We design a runtime that uses fast on-chip inter-core communication with small messages. At the same time, we use non coherent shared memory to avoid large core-to-core data transfers that would incur a high volume of unnecessary copying. We evaluate BDDT-SCC on a set of representative benchmarks, in terms of task granularity, locality, and communication. We find that memory locality and allocation plays a very important role in performance, as the architecture of the SCC memory controllers can create strong contention effects. We suggest patterns that improve memory locality and thus the performance of applications, and measure their impact.

cs.DC↗

arXiv2016

TwinCG: Dual Thread Redundancy with Forward Recovery for Conjugate Gradient Methods

Even though iterative solvers like the Conjugate Gradients method (CG) have been studied for over fifty years, fault tolerance for such solvers has seen much attention in recent years. For iterative solvers, two major reliable strategies of recovery exist: checkpoint-restart for backward recovery, or some type of redundancy technique for forward recovery. Important redundancy techniques like ABFT techniques for sparse matrix-vector products (SpMxV) have recently been proposed, which increase the resilience of CG methods. These techniques offer limited recovery options, and introduce a tolerable overhead. In this work, we study a more powerful resilience concept, which is redundant multithreading. It offers more generic and stronger recovery guarantees, including any soft faults in CG iterations (among others covering ABFT SpMxV), but also requires more resources. We carefully study this redundancy/efficiency conflict. We propose a fault tolerant CG method, called TwinCG, which introduces minimal wallclock time overhead, and significant advantages in detection and correction strategies. Our method uses Dual Modular Redundancy instead of the more expensive Triple Modular Redundancy; still, it retains the TMR advantages of fault correction. We describe, implement, and benchmark our iterative solver, and compare it in terms of efficiency and fault tolerance capabilities to state-of-the-art techniques. We find that before parallelization, TwinCG introduces around 5-6% runtime overhead compared to standard CG, and after parallelization efficiently uses BLAS. In the presence of faults, it reliably performs forward recovery for a range of problems, outperforming SpMxV ABFT solutions.

cs.DC↗

arXiv2016

Energy Optimization of Memory Intensive Parallel workloads

Energy consumption is an important concern in modern multicore processors. The energy consumed during the execution of an application can be minimized by tuning the hardware state utilizing knobs such as frequency, voltage etc. The existing theoretical work on energy mini- mization using Global DVFS (Dynamic Voltage and Frequency Scaling), despite being thorough, ignores the energy consumed by the CPU on memory accesses and the dynamic energy consumed by the idle cores. This article presents an analytical model for the performance and the overall energy consumed by the CPU chip on CPU instructions as well as the memory accesses without ignoring the dynamic energy consumed by the idle cores. We present an analytical framework around our energy-performance model to predict the operating frequencies for global DVFS that minimize the overall CPU energy consumption within a performance budget. Finally, we suggest a scheduling criteria for energy aware scheduling of memory intensive parallel applications.

cs.DC↗

arXiv2015

Evaluating Asymmetric Multicore Systems-on-Chip using Iso-Metrics

The end of Dennard scaling has pushed power consumption into a first order concern for current systems, on par with performance. As a result, near-threshold voltage computing (NTVC) has been proposed as a potential means to tackle the limited cooling capacity of CMOS technology. Hardware operating in NTV consumes significantly less power, at the cost of lower frequency, and thus reduced performance, as well as increased error rates. In this paper, we investigate if a low-power systems-on-chip, consisting of ARM's asymmetric big.LITTLE technology, can be an alternative to conventional high performance multicore processors in terms of power/energy in an unreliable scenario. For our study, we use the Conjugate Gradient solver, an algorithm representative of the computations performed by a large range of scientific and engineering codes.

cs.DC↗

arXiv2015

Iso-Quality of Service: Fairly Ranking Servers for Real-Time Data Analytics

We present a mathematically rigorous Quality-of-Service (QoS) metric which relates the achievable quality of service metric (QoS) for a real-time analytics service to the server energy cost of offering the service. Using a new iso-QoS evaluation methodology, we scale server resources to meet QoS targets and directly rank the servers in terms of their energy-efficiency and by extension cost of ownership. Our metric and method are platform-independent and enable fair comparison of datacenter compute servers with significant architectural diversity, including micro-servers. We deploy our metric and methodology to compare three servers running financial option pricing workloads on real-life market data. We find that server ranking is sensitive to data inputs and desired QoS level and that although scale-out micro-servers can be up to two times more energy-efficient than conventional heavyweight servers for the same target QoS, they are still six times less energy efficient than high-performance computational accelerators.

cs.DC↗