Identifying and localizing network problems using the PuNDIT project

21st International Conference on Computing in High Energy and Nuclear Physics

物理学;计算机科学

Batista, Jorge^1 ; Dovrolis, Constantine^2 ; Lee, Danny^2 ; McKee, Shawn^1

Randall Laboratory, Physics Department, University of Michigan, 450 Church Street, Ann Arbor

MI

48109-1040, United States^1

Klaus Advanced Computing Building, College of Computing, Georgia Institute of Technology, 266 Ferst Drive, Atlanta

GA

30332-0765, United States^2

关键词: De facto standard;    Distributed collaboration;    Effective infrastructures;    LHC computing grids;    Network diagnostics;    Network measurement infrastructure;    Open science grid;    Real-world networks;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/664/5/052027/pdf
DOI  :  10.1088/1742-6596/664/5/052027

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In today's world of distributed collaborations of scientists, there are many challenges to providing effective infrastructures to couple these groups of scientists with their shared computing and storage resources. The Pythia Network Diagnostic InfrasTructure (PuNDIT[1]) project is integrating and scaling research tools and creating robust code suitable for operational needs addressing the difficult challenge of automating the detection and location of network problems.PuNDIT is building upon the de-facto standard perfSONAR[2] network measurement infrastructure deployed in Open Science Grid(OSG)[3] and the Worldwide LHC Computing Grid(WLCG)[4]to gather and analyze complex real-world network topologies coupled with their corresponding network metrics to identify possible signatures of network problems from a set of symptoms. The PuNDIT Team is working closely with the perfSONAR developers from ESnet and Internet2 to integrate PuNDIT components as part of the perfSONAR Toolkit. A primary goal for PuNDIT is to convert complex network metrics into easily understood diagnoses in an automated way. We will report on the project progress to-date in working with the OSG and WLCG communities, describe the current implementation including some initial results and discuss future plans and the project timeline.