2026 ARC Winter Maintenance

2026 ARC Winter Maintenance

 

ARC Maintenance Updates

This page provides key details about the maintenance schedule and purpose.

Maintenance Dates

  • Great Lakes:
    • January 5-6
  • Armis2/Lighthouse:
    • January 5-8
  • Storage Globus Service:
    • Week of January 5th (no expected interruption)

HPC (High-Performance Computing) - All Clusters

Winter 2026

(version changes in bold)

Summer 2025

Red Hat 8.10

  • Kernel  4.18.0-553.85.1.el8_10.x86_64
  • glibc-2.28-251.el8_10.25.x86_64
  • ucx-1.16.0-1.2310409.x86_64 (OFED LTS provided)
  • gcc-8.5.0-28.el8_10.x86_64

Red Hat 8.10

  • Kernel  4.18.0-553.50.1.el8_10.x86_64
  • glibc-2.28-251.el8_10.16.x86_64
  • ucx-1.16.0-1.2310409.x86_64 (OFED LTS provided)
  • gcc-8.5.0-26.el8_10.x86_64

Mlnx-ofa_kernel-modules 

  • OFED-23.10-5.1.4

Mlnx-ofa_kernel-modules 

  • OFED-23.10-4.0.9

Slurm 25.11.1 compiled with:

  • PMIx
    • /opt/pmix/4.2.9
    • /opt/pmix/5.0.9
  • hwloc 2.2.0-3 (OS provided)
  • ucx-1.16.0-1.2310409.x86_64 (OFED LTS provided)
  • slurm-libpmi
  • slurm-contribs

Slurm 24.11.5 compiled with:

  • PMIx
    • /opt/pmix/4.2.9
    • /opt/pmix/5.0.7
  • hwloc 2.2.0-3 (OS provided)
  • ucx-1.16.0-1.2310409.x86_64 (OFED LTS provided)
  • slurm-libpmi
  • slurm-contribs

PMIx LD config /opt/pmix/4.2.9/lib

PMIx LD config /opt/pmix/4.2.9/lib

PMIx versions available in /opt:

  • 4.2.9
  • 5.0.9

PMIx versions available in /opt:

  • 4.2.9
  • 5.0.7

Singularity CE (Sylabs.io)

  • 4.1.5
  • 4.3.4

Singularity CE (Sylabs.io)

  • 4.1.3
  • 4.3.1

NVIDIA driver 580.105.08

  • CUDA 13_u2 support
    Might change back to 12.8 if older cards are not supported. 

NVIDIA driver 570.124.06

  • CUDA 12.8.1 support

Open OnDemand 4.0.8

 

Open OnDemand 4.0.1

System Changes:

  • Migrating to cgroup version 2 for both Slurm and future work.

  • To protect the stability of the login nodes, we’ve enabled a safeguard that can temporarily block new logins when a node is under heavy load.

    • If a login node is too busy, new SSH/login sessions may be refused.
    • In that case, your login attempt will fail and you’ll see a message explaining that the node is currently not accepting new sessions due to high resource usage.
    • Existing sessions and running jobs are not affected.
    • Once resource usage drops back to a safe level, new logins will be allowed again automatically.

    • If you repeatedly see this message for an extended period, please contact [email protected] with the time of your attempt and the node name, if known.

Slurm Release Notes: Slurm-25.11

New Features/Behaviors in Slurm 25.11:

  • Details to be added

HPC (High-Performance Computing) - Great Lakes System

  • New login nodes will be added during December

HPC (High-Performance Computing) - Armis2 and Lighthouse

The Data Center Engineering team will be working with contractors to perform the annual preventative maintenance on the data center where Armis2 and Lighthouse are racked.

HPC Software

Storage

  • The storage Globus nodes will be updated individually; the service will remain in production during these updates.

Globus

Version: To Be Determined (TBD)

SES (Service Environment Systems)

Contact Information

For questions or additional support, please contact ARC Support.

This format provides a clear breakdown of maintenance tasks and updates across different ARC services, with specifics on each section.