US ATLAS Computing Integration and Operations
-
-
13:00
→
13:15
Top of the Meeting 15mSpeakers: Eric Christian Lancon (BNL), Robert William Gardner Jr (University of Chicago (US))
-
13:15
→
13:20
ADC news and issues 5mSpeakers: Robert Ball (University of Michigan (US)), Wei Yang (SLAC National Accelerator Laboratory (US))
-
13:20
→
13:30
Production 10mSpeaker: Mark Sosebee (University of Texas at Arlington (US))
-
13:30
→
13:35
Data Management 5mSpeaker: Armen Vartapetian (University of Texas at Arlington (US))
-
13:35
→
13:40
Data transfers 5mSpeaker: Hironori Ito (Brookhaven National Laboratory (US))
-
13:40
→
13:45
Networks 5mSpeaker: Dr Shawn McKee (University of Michigan ATLAS Group)
Networking is part of the "Computing Models" component of the Community White Paper (CWP) and was discussed at the meeting in Annecy last week. Top level summary: the experiments (including ATLAS) seem to be happy with the current and planned work in networking. Near-term focus should be on increasing visibility into our networks and better enabling identification and localization of network problems. Longer term, it is desirable to work on network programmability and how the experiments may be able to benefit from software control/interaction with the network.
There are ongoing efforts to create publicly accessible dashboards showing network metrics, FTS data and transfer information from the LHCOPN and LHCONE networks. Already have a couple dashboards accessible at CERN: http://monit-grafana-open.cern.ch/?orgId=16
We need a campaign to clean up the perfSONAR instances, fixing problems with their updates and firewalls.
-
13:45
→
13:50
FAX and Xrootd Caching 5mSpeakers: Andrew Bohdan Hanushevsky (SLAC National Accelerator Laboratory (US)), Andrew Hanushevsky, Andrew Hanushevsky (STANFORD LINEAR ACCELERATOR CENTER), Ilija Vukotic (University of Chicago (US)), Wei Yang (SLAC National Accelerator Laboratory (US))
Addressing fine-grained authorization based on VO attributes.
Two rare bugs in Xrootd client's handling of metalink
Helping RAL to use xrootd proxy cache in front of their CEPH: Need a N2N to handle mapping between Object ID and storage path. Dealing with writing pass through in the cache, etc. - Work started.
-
13:50
→
14:00
OS performances testing 10mSpeaker: Doug Benjamin (Duke University (US))
-
14:00
→
14:15
HPCs integration 15mSpeaker: Taylor Childers (Argonne National Laboratory (US))
-
14:15
→
14:25
Singularity 10mSpeaker: Wei Yang (SLAC National Accelerator Laboratory (US))
-
14:25
→
16:00
Site Reports
- 14:25
-
14:30
AGLT2 5mSpeakers: Robert Ball (University of Michigan (US)), Dr Shawn McKee (University of Michigan ATLAS Group)
We have no known issues.
The MD3460/MD3060e disks did not appear to like being powered off for 3 days during the UM outage, there were 3 disk failures following power up. This is just something to keep in mind.
Auto-vacuum is not working on our dCache pgsql instances. We are still not sure that we have parameters correctly adjusted. The failure to vacuum caused a weekend outage that badly affected our June availability numbers. We have some advice from dCache support on how to modify the parameters, and will check to see if it is working, otherwise we'll cron up some discrete vacuuming.
-
14:35
MWT2 5mSpeakers: David Lesny (Univ. Illinois at Urbana-Champaign (US)), Lincoln Bryant (University of Chicago (US))
-
14:40
NET2 5mSpeaker: Prof. Saul Youssef (Boston University (US))
- 14:45
- 14:50
-
14:55
WT2 5mSpeaker: Wei Yang (SLAC National Accelerator Laboratory (US))
-
16:00
→
16:05
AOB 5m
-
13:00
→
13:15