2016 Open Science Grid all-hands meeting


Every spring, the entire Open Science Grid (OSG) community–consisting of resource owners and operators, users, and staff–gathers at the annual OSG all-hands meeting. The 2016 OSG all-hands meeting was held between Monday, March 14 and Thursday, March 17 at Clemson University in Clemson, SC, thanks in large part to Jim Bottum, the CIO and vice provost for technology at Clemson. The OSG is, as befitting a vehicle for distributed high-throughput computing, a highly distributed organization, with a community spread out across the US. As such, the all-hands meeting offers one of the few opportunities for face-to-face interaction for this community. Some of the highlights of the past year in the OSG were noted, including the passage of the 1 billion CPU hour/year threshold of production and the OSG’s role in providing part of the computational infrastructure for the LIGO experiment, which recently announced the observation of gravitational waves.

One considerable difference in the OSG landscape visible in this year’s OSG all-hands meeting as compared to the 2015 meeting held at Northwestern University was a complete lack of snow. Another was the considerable computing footprint that FIFE experiments have outside of Fermilab by way of OSG infrastructure and sites.

Opportunistic OSG usage by FIFE experiments in the 12-month period leading to the 2015 OSG AHM (top) and leading to the 2016 OSG AHM (bottom) shows a dramatic increase in usage.

Opportunistic OSG usage by FIFE experiments in the 12-month period leading to the 2015 OSG AHM (left) and leading to the 2016 OSG AHM (right) shows a dramatic increase in usage.

At the main plenary session on Wednesday, Ray Culbertson of Fermilab discussed the mu2e experiment and their rapid ramp-up in using the OSG [1]. A parallel session on Monday dedicated to non-LHC physics experiments saw presentations by Ken Herner about the DES follow-up-to-LIGO campaign [2], Steve Timm about DUNE [3], and a more technical discussion of mu2e’s OSG computing by Ray Culbertson [4]. Ken Herner also presented an overview of the FIFE project with an emphasis on recent progress using OSG resources at a plenary session on Tuesday [5].

Despite the “Grid” in Open Science Grid, recent years have shown the importance of getting OSG users access to “non-grid” resources such as commercial clouds and High Performance Computing (HPC) platforms. Burt Holzman presented progress on the Fermilab HEPCloud project, including successful use of Amazon Web Services resources by both CMS and NOvA [6]. Finally, the critical work of security in OSG operations continues; Jeny Teheran [7] and Dave Dykstra [8] reported on transitions to a federated identity system (CILogon).

1. The Mu2e Experiment at Fermilab : Experience with OSG Opportunistic
2. Dark Energy Survey on the OSG
3. Building DUNE on the Grid
4.The Mu2e Experiment at Fermilab : Second talk on Experience with OSG Opportunistic
5. FIFE Overview
6.The Fermilab HEPCloud
7.OSG CA Transition: Issues, challenges and lessons learned
8. Grid Access with Federated Identity

lunch-osg

Lunch at the OSG All-hands Meeting: Photo courtesy Ken Herner

— Bo Jayatilaka