Giuliano Casale http://wp.doc.ic.ac.uk/gcasale Tue, 07 Dec 2021 08:29:45 +0000 en-US hourly 1 2 papers accepted in IEEE MASCOTS http://wp.doc.ic.ac.uk/gcasale/2999-2/ Tue, 07 Dec 2021 08:28:44 +0000 https://wp.doc.ic.ac.uk/gcasale/?p=2999
]]>
Paper accepted in TPDS http://wp.doc.ic.ac.uk/gcasale/paper-accepted-in-tpds/ Thu, 10 Jun 2021 14:07:23 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2985 Read more »]]> We are happy to announce that our recent research work “COSCO: Container Orchestration using Co-Simulation and Gradient Based Optimization for Fog Computing Environments”, accepted for publication in IEEE Transactions on Parallel and Distributed Systems (https://ieeexplore.ieee.org/document/9448450). This work introduces a new framework that uses coupled simulation in tandem with gradients to inputs in a deep surrogate function, to facilitate QoS efficient decision making. Our code: https://github.com/imperial-qore/COSCO. An explanatory video is available here: https://www.youtube.com/watch?v=RZOWTj0rfBQ.

]]>
Paper accepted in IEEE/ACM TON http://wp.doc.ic.ac.uk/gcasale/new-paper-in-ieee-acm-ton/ Fri, 04 Dec 2020 05:18:58 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2968 Read more »]]> Our recent work Performance analysis of list-based caches with non-uniform access about stochastic modelling of caches has been accepted for publication in IEEE/ACM Transactions on Networking. This is a collaboration with Nicolas Gast (INRIA). The journal paper extends an earlier work published at INFOCOM 2018.

Abstract: List-based caches can offer lower miss rates than single-list caches, but their analysis is challenging due to state space explosion. In this setting, we propose novel methods to analyze performance for a general class of list-based caches with tree structure, non-uniform access to items and lists, and random or first-in first-out replacement policies. Even though the underlying Markov process is shown to admit a product-form solution, this is difficult to exploit for large caches. Thus, we develop novel approximations for cache performance metrics, in particular by means of a singular perturbation method and a refined mean field approximation. We compare the accuracy of these approaches to simulations, finding that our new methods rapidly converge to the equilibrium distribution as the number of items and the cache capacity grow in a fixed ratio. We find that
they are much more accurate than fixed point methods similar to prior work, with mean average errors typically below 1:5% even for very small caches. Our models are also generalized to account for synchronous requests, fetch latency, and item sizes, extending the applicability of approximations for list-based caches.

]]>
Vacancy: post-doc in real-time cloud monitoring data analysis http://wp.doc.ic.ac.uk/gcasale/vacancy-post-doc-in-real-time-cloud-monitoring-data-analysis/ Wed, 25 Nov 2020 13:38:05 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2964 Our group is seeking to fill a post-doctoral vacancy related to real-time anomaly detection and analysis of cloud monitor data. Applications are solicited from applicants with a relevant background in machine learning, real-time algorithms, or distributed systems. Applicants with a theoretical track-record in anomaly detection will also be considered.

]]>
New paper in MASCOTS (FaaS sizing) http://wp.doc.ic.ac.uk/gcasale/accepted-cold-start-aware-capacity-planning-for-faas-platforms/ Sat, 05 Sep 2020 14:37:41 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2902 Read more »]]> Our recent work COCOA: Cold Start Aware Capacity Planning for Function-as-a-Service Platforms about modelling and sizing FaaS platforms taking into account cold-starts has been accepted in IEEE MASCOTS 2020, congratulations to Alim Ul Gias!

Abstract: Function-as-a-Service (FaaS) is increasingly popular in the software industry due to the implied cost-savings in event-driven workloads and its synergy with DevOps. To size an on-premise FaaS platform, it is important to estimate the required CPU and memory capacity to serve the expected loads. Given the service-level agreements, it is however challenging to take the cold start issue into account during the sizing process. We have investigated the similarity of this problem with the hit rate improvement problem in TTL caches and concluded that solutions for TTL cache, although potentially applicable, lead to over-provisioning in FaaS. Thus, we propose a novel approach, COCOA, to solve this issue. COCOA uses a queueing-based approach to assess the effect of cold starts on FaaS response times. It also considers different memory consumption values depending on whether the function is idle or in execution. Using an event-driven FaaS simulator, FaasSim, we have developed, we show that COCOA can reduce over-provisioning by over 70% in some workloads, while satisfying the service-level agreements.

]]>
Paper accepted in IEEE Access (Spark anomaly detection) http://wp.doc.ic.ac.uk/gcasale/accepted-anomaly-detection-fo-data-streaming-systems-ieee-access/ Fri, 31 Jul 2020 14:31:00 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2895 Read more »]]> Congratulations to Ahmad Alnafessah for our recently accepted paper TRACK-Plus: Optimizing Artificial Neural Networks for Hybrid Anomaly Detection in Data Streaming ! The paper is published in IEEE Access.

Abstract: Software applications can feature intrinsic variability in their execution time due to interference from other applications or software contention from other users, which may lead to unexpectedly long running times and anomalous performance. There is thus a need for effective automated performance anomaly detection methods that can be used within production environments to avoid any late detection of unexpected degradations of service level. To address this challenge, we introduce TRACK-Plus a black-box training methodology for performance anomaly detection. The method uses an artificial neural networks-driven methodology and Bayesian Optimization to identify anomalous performance and are validated on Apache Spark Streaming. TRACK-Plus has been extensively validated using a real Apache Spark Streaming system and achieve a high F-score while simultaneously reducing training time by 80% compared to efficiently detect anomalies.

]]>
New paper in Winter Sim. Conf. (LINE 2.0) http://wp.doc.ic.ac.uk/gcasale/paper-accepted-in-wsc/ Mon, 15 Jun 2020 06:06:14 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2882 Our latest paper “Integrated performance evaluation of extended queueing network models with LINE” has now been accepted at the Winter Simulation conference 2020, the flagship event of SIGSIM. The paper is the first one to present in details our LINE 2.0 solver.

]]>
Paper accepted in ITL (iThermoFog) http://wp.doc.ic.ac.uk/gcasale/paper-accepted-in-itl/ Mon, 15 Jun 2020 06:03:03 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2877 Read more »]]> The manuscript “iThermoFog: IoT-Fog based Automatic Thermal Profile Creation for Cloud Data Centers using Artificial Intelligence Techniques” by Shreshth Tuli, Sukpahl Gill, Giuliano Casale, and Nick Jennings, has been accepted in Internet Technology Letters. The paper presents iThermoFog, which develops an AI based automatic model for creating thermal profiles in Cloud Data centers used as backends for IoT/Fog applications.

]]>
New paper in PEVA (Fluid DPS) http://wp.doc.ic.ac.uk/gcasale/paper-accepted-in-elsevier-peva/ Fri, 06 Mar 2020 10:43:44 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2853 Our work on modelling discriminatory processor sharing in closed queueing networks has been accepted in Elsevier Performance Evaluation, congratulations to Lulai Zhu and Iker Perez for the great collaboration!

]]>
Paper accepted in Cluster Computing http://wp.doc.ic.ac.uk/gcasale/paper-accepted-in-cluster-computing/ Sun, 27 Oct 2019 11:47:32 +0000 http://wp.doc.ic.ac.uk/gcasale/?p=2831 Our work on automated anomaly detection for Apache Spark has been accepted in Springer Cluster Computing, congratulations to Ahmad Alnafessah for leading this work!

]]>