Conference paper Open Access

Contextual linking between workflow provenance and system performance logs

Ahanach, Elias el Khaldi; Koulouzis, Spiros; Zhao, Zhiming

Citation Style Language JSON Export

  "DOI": "10.1109/eScience.2019.00093", 
  "title": "Contextual linking between workflow provenance and system performance logs", 
  "issued": {
    "date-parts": [
  "abstract": "<p>When executing scientific workflows, anomalies of</p>\n\n<p>the workflow behavior are often caused by different issues such</p>\n\n<p>as resource failures at the underlying infrastructure. The provenance</p>\n\n<p>information collected by workflow management systems</p>\n\n<p>only captures the transformation of data at the workflow level.</p>\n\n<p>Analyzing provenance information and apposite system metrics</p>\n\n<p>requires expertise and manual effort. Moreover, it is often timeconsuming</p>\n\n<p>to aggregate this information and correlate events</p>\n\n<p>occurring at different levels of the infrastructure. In this paper,</p>\n\n<p>we propose an architecture to automate the integration among</p>\n\n<p>workflow provenance information and performance information</p>\n\n<p>from the infrastructure level. Our architecture enables workflow</p>\n\n<p>developers or domain scientists to effectively browse workflow</p>\n\n<p>execution information together with the system metrics, and</p>\n\n<p>analyze contextual information for possible anomalies.</p>", 
  "author": [
      "family": "Ahanach, Elias el Khaldi"
      "family": "Koulouzis, Spiros"
      "family": "Zhao, Zhiming"
  "id": "3462820", 
  "event-place": "San Diego", 
  "version": "Camera ready", 
  "type": "paper-conference", 
  "event": "IEEE International Conference on eScience 2019 (eScience2019)"
Views 272
Downloads 97
Data volume 29.8 MB
Unique views 270
Unique downloads 95


Cite as