Abstract
The Linux kernel feature Control Groups (cgroups) has been used to gather metrics on the resource usage of single and eight-core ATLAS workloads. It has been used to study the effects on performance of a reduction in the amount of physical memory. The results were used to optimise cluster performance, and consequently increase cluster throughput by up to 10%.
Highlights
Beginning in 2015, the ATLAS experiment started to deliver eight-core production workloads to the sites of the WLCG
In the case of analysis, this over-estimation is a function of the payloads executed at UKI-SCOTGRID-GLASGOW and may not be indicative of all other sites
There are two changes that ATLAS could make to improve utilisation of system resources. If they were to distinguish between eight-core simulation and eight-core reconstruction jobs at the queue or pilot level, an improvement in utilisation of up to 31% could be obtained
Summary
Beginning in 2015, the ATLAS experiment started to deliver eight-core production workloads to the sites of the WLCG. To gain a deeper understanding of cgroups and to better understand the performance of ATLAS eight-core simulation payloads, a study was carried out which capped the available physical memory of jobs. This allowed the performance and behaviour of such jobs in memorystarved environments to be examined. From data collected between January and December 2015, memory and CPU usage footprints were extracted and the type of ATLAS[3] workload was identified This data has been used to optimise the amount of memory allocated to ATLAS workloads (initially ATLAS single-core production workloads) in order to increase cluster utilisation by up to 10%. Published under licence by IOP Publishing Ltd doi:10.1088/1742-6596/762/1/012010
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
