Web System for Data Quality Assessment of Tile Calorimeter During the ATLAS Operation

TileCal is the barrel hadronic calorimeter of the ATLAS experiment and has ~10 000 electronic channels. Supervising the detector behavior is a very important task to ensure proper operation. Collaborators perform analyzes over reconstructed data of calibration runs in order to give detailed considerations about failures and to assert the equipment status. Then, the data quality responsible provides the list of problematic channels that should not be considered for physics analysis. Since the commissioning period, our group has developed seven web systems that guide the collaborators through the data quality assessment task. Each system covers a part of the job, providing information on the latest runs, displaying status from the automatic monitoring framework, giving details about power supplies operation, presenting the generated plots and storing the validation outcomes, assisting to write logbook entries, creating and submitting the bad channels list to the conditions database and publishing the equipment performance history. Due to the beginning of the operation, runs are acquired more often. The increasing amount of data represents a new challenge to the systems. Until now, over ninety million plots are stored in the databases, affecting the performance of the systems. Thus, an optimized data model was designed to reduce the number of needed queries. The web systems were reassembled in a unique system to provide an integra ted view of the validating process. The server load was minimized by using asynchronous requests from the browser. This article describes the new approach and reports the performance improvement.v