Presentation
(PP03): NeIC presents Nordic Project Highlights: Nordic Resource Exchange & Highly-Available Storage Within Nordic LHC Tier-1
SessionProject Poster Presentation
Event Type
Project Poster
Big Data Analytics
Clouds and Distributed Computing
Reliability
Resiliency
Storage Technologies
TimeWednesday, June 19th3:15pm - 4pm
LocationBooth N-230
DescriptionThrough the Nordic e-Infrastructure Collaboration (NeIC) together the Nordic countries are tackling e-infrastructure challenges beyond singular national capabilities. Specific aspects of two current NeIC projects are highlighted on this poster.
Dellingr, NeIC's cross-border resource sharing project, contributes towards establishing a framework for resource sharing across the Nordics. Based on the experience of the first pilot the intented improved resource access process is presented. Allocations will use billing units (BU) as a common currency and the Waldur cloud brokerage framework (https://share.neic.no) will be the central tool used throughout the whole process starting from resource application, over usage tracking to linking to resulting publications. https://dellingr.neic.no/apply/
Within the context of the Nordic Large Hadron Collider (LHC) Computing Tier-1 the effective high-availability routines since 2016 are presented.
Conceived in 2002 as a one-of-a-kind facility unifying six academic computing centres in four different countries, the Nordic Tier-1 can now look back on more than 13 years of successful operations with an excellent track record on reliability and availability. Using the High-Availability (HA) features within dCache, ucarpd for automatic IP fail-over, HAProxy for load balancing and Ganeti as well as repmgr for managing VMs and HA PostgreSQL, no downtimes have to be scheduled for regular Linux security updates or dCache upgrades and no restrictions for the perceived user experience apply.
Dellingr, NeIC's cross-border resource sharing project, contributes towards establishing a framework for resource sharing across the Nordics. Based on the experience of the first pilot the intented improved resource access process is presented. Allocations will use billing units (BU) as a common currency and the Waldur cloud brokerage framework (https://share.neic.no) will be the central tool used throughout the whole process starting from resource application, over usage tracking to linking to resulting publications. https://dellingr.neic.no/apply/
Within the context of the Nordic Large Hadron Collider (LHC) Computing Tier-1 the effective high-availability routines since 2016 are presented.
Conceived in 2002 as a one-of-a-kind facility unifying six academic computing centres in four different countries, the Nordic Tier-1 can now look back on more than 13 years of successful operations with an excellent track record on reliability and availability. Using the High-Availability (HA) features within dCache, ucarpd for automatic IP fail-over, HAProxy for load balancing and Ganeti as well as repmgr for managing VMs and HA PostgreSQL, no downtimes have to be scheduled for regular Linux security updates or dCache upgrades and no restrictions for the perceived user experience apply.
Poster PDF
Poster Author
