NOTE to visitors to the Michelson Center for Physics: enter through south doors.
A live hackathon is being organized to develop a versatile hybrid Kubernetes platform tailored for Facility R&D endeavors. This platform will cater to scalable, low-latency analysis systems, alongside providing persistent infrastructure for notebook-based training sessions focused on ATLAS software tutorials, machine learning, and cutting-edge analysis tools designed for HL-LHC.
The meeting will be hosted by the Enrico Fermi Institute at the University of Chicago. Support is provided by the US ATLAS Operations program and the IRIS-HEP Scalable Systems Laboratory team and will be 100% hands-on and in-person: no presentations or remote participation (except for training on the first day). Participants are encouraged to come prepared with their own K8s resource with Kubernetes 1.29.2 installed.
The concrete goal will be to run a variety of scalable multi-cluster demonstrators:
- A low-latency, Run 3 type analysis (driven from a notebook)
- Advanced workflows utilizing the Uproot/Awkward/Dask backend with PHYSLITE
- Advanced workflows utilizing RDataFrame
On the first morning we will hold tutorial / training sessions for K8s cluster admins. Attendance for this portion can be remote.
