Cargando…
Exploring improvements to the LHCb ELOG electronic logbook
ELOG, the electronic logbook used by LHCb, has been suffering from poor accessibility and slow performance. Investigating revealed that a combination of inefficient searches, caused mainly by frequent reads of files stored in NFS, and an inability for the server to handle concurrent operations were...
Autores principales: | , |
---|---|
Lenguaje: | eng |
Publicado: |
2015
|
Materias: | |
Acceso en línea: | http://cds.cern.ch/record/2048906 |
Sumario: | ELOG, the electronic logbook used by LHCb, has been suffering from poor accessibility and slow performance. Investigating revealed that a combination of inefficient searches, caused mainly by frequent reads of files stored in NFS, and an inability for the server to handle concurrent operations were rendering the service unusable for up to a minute when users performed search or sort operations. By adding a minor patch to the ELOG source code, moving data to local storage and optimizing server configuration the project was able to reduce search-times for the largest of the logbooks being used at LHCb to 30\% while also improving possibilities for future growth by allowing for concurrent use and accelerating the most common search-operations in a way which should stay consistent over extended time-periods. |
---|