• nsdi2020

    Accepted author manuscript, 657 KB, PDF-document

  • Alexandru Uta
  • Alexandru Custura
  • Dmitry Duplyakin
  • Ivo Jimenez
  • Jan S. Rellermeyer
  • Carlos Maltzahn
  • Robert Ricci
  • Alexandru Iosup
Performance variability has been acknowledged as a problem for over a decade by cloud practitioners and performance engineers. Yet, our survey of top systems conferences reveals that the research community regularly disregards variability when running experiments in the cloud. Focusing on networks, we assess the impact of variability on big-data cloud-based workloads by gathering large-scale traces from mainstream commercial clouds and private research clouds. Our data collection consists of millions of datapoints gathered while transferring over 9 petabytes of data. We characterize the network variability present in our data and show that, even though commercial cloud providers implement mechanisms for quality-of-service enforcement, variability still occurs, and is even exacerbated by such mechanisms and service provider policies. We show how big-data workloads suffer from significant slowdowns and lack predictability and replicability, even when state-of-the-art experimentation techniques are used. We provide guidelines for practitioners to reduce the volatility of big data performance, making experiments more repeatable.
Original languageEnglish
Title of host publicationProceedings of the 17th USENIX Symposium on Networked Systems Design and Implementation
Number of pages15
Publication statusPublished - 25 Feb 2020
Event17th USENIX Symposium on Networked Systems Design and Implementation - Santa Clara, United States
Duration: 25 Feb 202027 Feb 2020
Conference number: 2020


Conference17th USENIX Symposium on Networked Systems Design and Implementation
Abbreviated titleNSDI
CountryUnited States
CitySanta Clara
Internet address

ID: 67420026