Cargando…

EOS deployment at GRIF

<!--HTML-->In this communication, we are going to present the deployment project of the EOS storage software solution at the GRIF site. GRIF is a distributed site made of four (4) different subsites, in different locations of the Paris region. The worst network latency between the subsites is...

Descripción completa

Detalles Bibliográficos
Autor principal: Vamvakopoulos, Emmanouil
Lenguaje:eng
Publicado: 2022
Materias:
Acceso en línea:http://cds.cern.ch/record/2803631
_version_ 1780972803120431104
author Vamvakopoulos, Emmanouil
author_facet Vamvakopoulos, Emmanouil
author_sort Vamvakopoulos, Emmanouil
collection CERN
description <!--HTML-->In this communication, we are going to present the deployment project of the EOS storage software solution at the GRIF site. GRIF is a distributed site made of four (4) different subsites, in different locations of the Paris region. The worst network latency between the subsites is within 2-4 msec with 3 of them connected with a 100G connection. The objective is to consolidate the four (4) currently independent DPM instances into (a new) one EOS instance. The EOS capabilities such as the service structure (simple roles based on xrootd), the failover mechanism, the redundancy of metadata nodes, and the backend key/store database (quarkdb)) ensure with a straightforward installation and configuration the high availability requirements for the geographically distributed environment of GRIF. We are going to discuss the future EOS service structure and capacities at GRIF and present a summary of the organization of the EOS filesystems over legacy raid6 storage devices on heterogeneous hardware.
id cern-2803631
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2022
record_format invenio
spelling cern-28036312022-11-02T22:03:51Zhttp://cds.cern.ch/record/2803631engVamvakopoulos, EmmanouilEOS deployment at GRIFEOS workshopHEP Computing<!--HTML-->In this communication, we are going to present the deployment project of the EOS storage software solution at the GRIF site. GRIF is a distributed site made of four (4) different subsites, in different locations of the Paris region. The worst network latency between the subsites is within 2-4 msec with 3 of them connected with a 100G connection. The objective is to consolidate the four (4) currently independent DPM instances into (a new) one EOS instance. The EOS capabilities such as the service structure (simple roles based on xrootd), the failover mechanism, the redundancy of metadata nodes, and the backend key/store database (quarkdb)) ensure with a straightforward installation and configuration the high availability requirements for the geographically distributed environment of GRIF. We are going to discuss the future EOS service structure and capacities at GRIF and present a summary of the organization of the EOS filesystems over legacy raid6 storage devices on heterogeneous hardware.oai:cds.cern.ch:28036312022
spellingShingle HEP Computing
Vamvakopoulos, Emmanouil
EOS deployment at GRIF
title EOS deployment at GRIF
title_full EOS deployment at GRIF
title_fullStr EOS deployment at GRIF
title_full_unstemmed EOS deployment at GRIF
title_short EOS deployment at GRIF
title_sort eos deployment at grif
topic HEP Computing
url http://cds.cern.ch/record/2803631
work_keys_str_mv AT vamvakopoulosemmanouil eosdeploymentatgrif
AT vamvakopoulosemmanouil eosworkshop