FAS Research Computing - Cluster CVMFS issue causing node closure – Incident details

All systems operational

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.

Cluster CVMFS issue causing node closure

Resolved
Degraded performance
Started 22 days agoLasted 5 days

Affected

Cannon Cluster

Degraded performance from 1:40 PM to 4:31 AM

SLURM Scheduler - Cannon

Degraded performance from 1:40 PM to 4:31 AM

Cannon Compute Cluster (Holyoke)

Degraded performance from 1:40 PM to 4:31 AM

Boston Compute Nodes

Degraded performance from 1:40 PM to 4:31 AM

GPU nodes (Holyoke)

Degraded performance from 1:40 PM to 4:31 AM

seas_compute

Degraded performance from 1:40 PM to 4:31 AM

Updates
  • Resolved
    UTC
    Resolved

    This incident has been resolved. Less than 18 nodes remain in this state and will be cleaned up soon.

  • Monitoring
    UTC
    Monitoring

    We have rolled out a fix and are currently monitoring the result.

    Any nodes still in Kill Task Fail will need to fully drain of running jobs first before being reopened to ensure there are no orphaned processes.

    PLEASE NOTE: This will take quite some time to fully resolve as jobs are still running on these nodes. But the nodes cannot accept new jobs until cleared.

    Currently # of potentially impacted nodes: 18 (of an original 600 - number will update periodically)

  • Investigating
    UTC
    Investigating

    We are currently investigating this incident. Due to a CVMFS mount issue, some nodes are being closed and labelled as "Kill task fail"
    WIP

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