ARC-TS Summer 2020 Maintenance

 

Due to maintenance, the high-performance computing (HPC) clusters and their storage systems will be unavailable:

  • August 3, 2020, 7 a.m. to 5 p.m.: Great Lakes HPC Cluster

  • August 4, 2020, 7 a.m. to 5 p.m.: Armis2 and Lighthouse HPC Clusters

  • August 3, 2020, 8 a.m. to August 4, 2020, 5 p.m.:  ConFlux HPC Cluster

Planned updates include:

Great Lakes:

  • Migration to CentOS 7.8
  • Upgrade NVIDIA drivers for CUDA 10.2
  • Upgrade Slurm to 20.02.3
  • Upgrade to latest OFED
  • Update proxy service to newer hardware and OS
  • Updates to GPFS filesystem for /scratch
  • Updates to the InfiniBand networking (firmware on switches and HCAs)
  • Update software defaults for a variety of our software packages
  • Potentially updating to Open OnDemand 1.7.1

Lighthouse:

  • Migration to CentOS 7.8
  • Upgrade NVIDIA drivers for CUDA 10.2
  • Upgrade Slurm to 20.x, pending testing
  • Upgrade to latest OFED
  • Update proxy service to newer hardware and OS
  • Updates to the InfiniBand networking (firmware on switches and HCAs)
  • Update software defaults for a variety of our software packages
  • Potentially updating to Open OnDemand 1.7.1
  • Move IPoIB to new, larger (/23 TBD) subnet

Armis2:

  • Migration to CentOS 7.8
  • Upgrade NVIDIA drivers for CUDA 10.2
  • Upgrade Slurm to 20.x, pending testing
  • Upgrade to latest OFED
  • Update proxy service to newer hardware and OS
  • Updates to the InfiniBand networking (firmware on switches and HCAs)
  • Update software defaults for a variety of our software packages