EOS Monitoring and Analytics Tools

Project Specification

The IT DSS group at CERN runs and evaluates innovative cloud storage technologies for their application to big data problems in high-energy physics research. One of the main storage systems is EOS, a multi petabyte disk storage built from commodity hardware heavily used by LHC and non-LHC experiments, primarily for physics data. In the scope of a common monitoring framework for our cloud storage services and in order to improve the analytics, the manageability and the user interface, this project’s aim was to:

•  implement a probing system to check storage health and main KPI
• implement features for the web visualization of storage KPIs
• investigate the integration of other monitoring within the framework
• compare different solution of storage monitoring and analytic tools

Abstract

The aim of this openlab project was to improve user experience of EOS, a disk based data storage system used by LHC and non-LHC users. The online monitoring systems for EOS are various and this project main focus was on improving EOS Cockpit structure and functionalities and simplifying SLS probing system. The EOS Cockpit structure was changes according to users’ requirements, and new features were implemented. The SLS probing script was significantly simplified and adapted to the next generation tools used for displaying service levels.