FAS Research Computing - Bemærk historie

Oplever delvist forringet ydeevne

Status page for the Harvard FAS Research Computing cluster and other resources.

Cluster Utilization (VPN and FASRC login required): Cannon | FASSE


Please scroll down to see details on any Incidents or maintenance notices.
Monthly maintenance occurs on the first Monday of the month (except holidays).

GETTING HELP
Documentation: https://docs.rc.fas.harvard.edu | Account Portal https://portal.rc.fas.harvard.edu
Email: rchelp@rc.fas.harvard.edu | Support Hours

The colors shown in the bars below were chosen to increase visibility for color-blind visitors.
For higher contrast, switch to light mode at the bottom of this page if the background is dark and colors are muted.
Begrænset drift

SLURM Scheduler - Cannon - Begrænset drift

Cannon Compute Cluster (Holyoke) - Begrænset drift

Boston Compute Nodes - Begrænset drift

GPU nodes (Holyoke) - Begrænset drift

seas_compute - Begrænset drift

Operationel

SLURM Scheduler - FASSE - Operationel

FASSE Compute Cluster (Holyoke) - Operationel

Operationel

Kempner Cluster CPU - Operationel

Kempner Cluster GPU - Operationel

Operationel

FASSE login nodes - Operationel

Operationel

Cannon Open OnDemand/VDI - Operationel

FASSE Open OnDemand/VDI - Operationel

Operationel

Netscratch (Global Scratch) - Operationel

Home Directory Storage - Boston - Operationel

Tape - (Tier 3) - Operationel

Holylabs - Operationel

Isilon Storage Holyoke (Tier 1) - Operationel

Holystore01 (Tier 0) - Operationel

HolyLFS04 (Tier 0) - Operationel

HolyLFS05 (Tier 0) - Operationel

HolyLFS06 (Tier 0) - Operationel

Holyoke Tier 2 NFS (new) - Operationel

Holyoke Specialty Storage - Operationel

holECS - Operationel

Isilon Storage Boston (Tier 1) - Operationel

BosLFS02 (Tier 0) - Operationel

Boston Tier 2 NFS (new) - Operationel

CEPH Storage Boston (Tier 2) - Operationel

Boston Specialty Storage - Operationel

bosECS - Operationel

Samba Cluster - Operationel

Globus Data Transfer - Operationel

Bemærk historie

dec. 2025

Scheduling down on Cannon
  • Løst
    Løst

    We have found the malformed job which caused the scheduler to wedge and have cancelled it. We will continue to work with SchedMD to prevent future stalls. The scheduler is back in operation. Happy Holidays!

  • Opdatering
    Opdatering

    We have filed a ticket with SchedMD to diagnose the issue (https://support.schedmd.com/show_bug.cgi?id=24376). All partitions are set to DOWN as we assess the situation

  • Undersøger
    Undersøger

    The slurm scheduler ground to halt at 3pm EST. We are investigating the cause.

Bauer MiniLIMS is down
  • Løst
    Løst
    Bauer MiniLIMS is back up. This incident was automatically resolved.
  • Undersøger
    Undersøger
    Bauer MiniLIMS is down at the moment. This incident was automatically created.
Login node reboots
  • Løst
    Løst

    Login nodes have been rebooted, and tier 1 shares are accessible from there.

    This incident has been resolved.

  • Undersøger
    Undersøger

    Due to Tier 1 Isilon shares not mounting properly, login nodes will need to be rebooted.

    We are currently investigating this incident.

Isilon shares not accessible
  • Løst
    Løst

    This incident has been resolved. All Tier 1 shares are accessible.

  • Undersøger
    Undersøger

    Storage shares on Tier 1 Isilon may not be accessible at this time. We are currently investigating this incident.

Authentication outage
  • Løst
    Løst

    Openauth/radius is now operational. This update was created by an automated monitoring service.

  • Undersøger
    Undersøger

    Authentication issues with openauth/radius. This incident was created by an automated monitoring service.

nov. 2025

Coldfront outage
  • Løst
    Løst

    This alert was a false alarm. Coldfront is operating normally.

  • Undersøger
    Undersøger
    Coldfront cannot be accessed at the moment. This incident was created by an automated monitoring service.
Authentication outage
  • Løst
    Løst

    This test incident has been resolved.

  • Undersøger
    Undersøger

    Authentication issues with openauth/radius. This incident was created by an automated monitoring service.

Slurm scheduler unplanned maintenance
  • Løst
    Løst
    The Slurm scheduler is back online.
  • Identificeret
    Identificeret
    The Slurm scheduler is experiencing a problem and we need to take it down briefly to clear this up. Should be short duration, but no ETA yet. This will not affect running jobs but will mean you cannot talk to the scheduler or submit new jobs.
Starfish dashboard maintenance Nov. 14th 5-6PM
  • Gennemført
    november 14, 2025 ved 23:00
    Gennemført
    november 14, 2025 ved 23:00
    Maintenance has completed successfully
  • Igangværende
    november 14, 2025 ved 22:00
    Igangværende
    november 14, 2025 ved 22:00
    Maintenance is now in progress
  • Planlagt
    november 14, 2025 ved 22:00
    Planlagt
    november 14, 2025 ved 22:00

    There is a planned upgrade of the Starfish dashboard scheduled for Friday November 14th starting at 5PM.
    The dashboard will be down for an hour while the upgrade is performed.

Slurm scheduler issues - WIP
  • Løst
    Løst

    The Slurm scheduler is back up and new jobs can be scheduled at this time.

    This incident has been resolved.

  • Identificeret
    Identificeret
    The Slurm job scheduler will be offline temporarily while we resolve some minor issues. Jobs will continue to run but you will be unable to access the scheduler or submit new jobs until we return it to service.

okt. 2025

Holystore01 is degraded.
  • Løst
    Løst
    This incident has been resolved.
  • Undersøger
    Undersøger

    We are currently investigating an issue with holystore01. Service may be degraded. Updates will be posted here as we identify the root cause.

License servers down
  • Løst
    Løst

    The license server is responding normally again.

  • Opdatering
    Opdatering
    We are currently investigating this incident.
  • Undersøger
    Undersøger

    rclic is currently down. Software licenses may not be accessible, particularly for the following:

    • Matlab

    • Gurobi

    • Mathematica

    We are currently investigating this incident.

FASRC monthly maintenance Monday October 6th, 2025 9am-1pm
  • Gennemført
    oktober 06, 2025 ved 17:00
    Gennemført
    oktober 06, 2025 ved 17:00
    Maintenance is now in progress
  • Igangværende
    oktober 06, 2025 ved 13:00
    Igangværende
    oktober 06, 2025 ved 13:00
    Maintenance is now in progress
  • Planlagt
    oktober 06, 2025 ved 13:00
    Planlagt
    oktober 06, 2025 ved 13:00

    FASRC monthly maintenance will take place Monday October 6th, 2025 from 9am-1pm

    NOTICES

    MAINTENANCE TASKS
    Cannon cluster will be paused during this maintenance?: NO
    FASSE cluster will be paused during this maintenance?: NO

    • DNS server reboots

      • Audience: All FASRC services

      • Impact: Rolling reboot should have no impact

    • Login node reboots

      • Audience: Anyone logged into a FASRC Cannon or FASSE login node

      • Impact: All login nodes will rebooted during this maintenance window

    • Netscratch cleanup ( https://docs.rc.fas.harvard.edu/kb/policy-scratch/ )

      • Audience: Cluster users

      • Impact: Files older than 90 days will be removed. Please note that retention cleanup can and does run at any time, not just during the maintenance window.

    Thank you,
    FAS Research Computing
    https://docs.rc.fas.harvard.edu/
    https://www.rc.fas.harvard.edu/

OOD, FASRC authentication down
  • Løst
    Løst

    This incident has been resolved and login to FASRC sites and services, including OOD and Spinal, is restored.

  • Identificeret
    Identificeret

    Some FASRC webpages and services (such as OOD) that require authentication may not load.

    We have identified the cause and are working on a solution

okt. 2025 til dec. 2025

Næste

For issues not shown here, please contact FASRC via
https://portal.rc.fas.harvard.edu or email rchelp@rc.fas.harvard.edu