Weekly meeting to co-ordinate DUNE computing workflow and workload development work, at 8am CT.
The 5 minute time allocations are just placeholders and the names next to each topic are incomplete!
https://fnal.zoom.us/j/94131720980
Meeting ID: 941 3172 0980
Usual computing password (same as Monday general computing meetings)
Issue tracker: https://github.com/DUNE/dist-comp/labels/WMSdev
Wiki: https://wiki.dunescience.org/wiki/Workload_Management
Steve Timm, Ken Herner, and I had a short discussion. Mainly about the DUNE Global Pool.
- there was a request for a network variance to allow pilots to connect back to FNAL from offsite so that they could join the HTCondor pool. The variance was approved, but isn't yet working. Steve will continue to try and probe the port via telnet from CERN, and inform others when he is successful.
- the question of how we're going to authenticate the offsite submit node in the time before tokens. Steve commented that GSI isn't long for this world, so might not be worth the effort to configure. "FS" (Kirby believes this was common file system) was an option, but not sure if it's the right way to go. A topic to discuss
- Kirby hasn't gotten a response from anyone at BNL willing to help setup a submit node there for testing. Will continue to try and get a point of contact to work on this. Setting up a submit node at CERN using a VM that Steve Timm can setup is a likely second option
- jobsub lite will be invited to give an update at the next workflow/load development meeting
- meeting was gaveled to adjournment without opposition at 8:19 am CT
Including the DUNE Global Pool
Including jobsub
Including interaction with Data Management
Where relevant to workload: benchmarking, ETF monitoring of job execution, CRIC