Speaker
Description
URL for further information
https://twiki.cern.ch/twiki/bin/view/CMS/PADASiteCommissioning
Impact
The foreseen impact of this work is to significantly improve the awareness of problems at CMS sites preventing CMS workloads from being smoothly executed. This translates to a better performance of the whole CMS computing system, a more efficient usage of the resources and less work for the operators and the physicists using the Grid.
Keywords
CMS, Commissioning, Sites, Reliability, Grid, Monitoring
Conclusions and Future Work
The work presented here is ongoing, constantly adapting to the needs of the CMS computing and subject to constant improvements.
Detailed analysis
The CMS computing system relies on resources distributed all over the world, including a Tier-0 at CERN, seven Tier-1 sites and around 50 Tier-2 sites, belonging to three different Grids, EGEE, OSG and NorduGrid. The reliability of these sites needs to be measured for several reasons: to plan organized simulation or reprocessing activities or define which sites are available for user analysis, to expose all possible problems relevant for CMS and to help in solving them. The area covered by the Grid operations does not include some functionality specific to CMS: therefore new tools have been developed in CMS to expand the site testing in relation with CMS-specific aspects and to measure the site performance in a way which is directly related to the CMS use cases. In addition to that, a web interface was developed to collect all the relevant information and define as simply as possible if a site can be considered to be "usable" for CMS workloads.
