FAS Research Computing - Historique des incidents

Holyoke Tier 2 NFS connaît une panne partielle

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.
Opérationnel

SLURM Scheduler - Cannon - Opérationnel

Cannon Compute Cluster (Holyoke) - Opérationnel

Boston Compute Nodes - Opérationnel

GPU nodes (Holyoke) - Opérationnel

seas_compute - Opérationnel

Opérationnel

SLURM Scheduler - FASSE - Opérationnel

FASSE Compute Cluster (Holyoke) - Opérationnel

Opérationnel

Kempner Cluster CPU - Opérationnel

Kempner Cluster GPU - Opérationnel

Opérationnel

FASSE login nodes - Opérationnel

Opérationnel

Cannon Open OnDemand - Opérationnel

FASSE Open OnDemand - Opérationnel

Panne partielle

Netscratch (Global Scratch) - Opérationnel

Home Directory Storage - Boston - Opérationnel

Tape - (Tier 3) - Opérationnel

Holylabs - Opérationnel

Isilon Storage Holyoke (Tier 1) - Opérationnel

Holystore01 (Tier 0) - Opérationnel

HolyLFS04 (Tier 0) - Opérationnel

HolyLFS05 (Tier 0) - Opérationnel

HolyLFS06 (Tier 0) - Opérationnel

Holyoke Tier 2 NFS - Panne partielle

Holyoke Specialty Storage - Opérationnel

holECS - Opérationnel

Isilon Storage Boston (Tier 1) - Opérationnel

BosLFS02 (Tier 0) - Opérationnel

Boston Tier 2 NFS - Opérationnel

CEPH Storage Boston (Tier 2) - Opérationnel

Boston Specialty Storage - Opérationnel

bosECS - Opérationnel

Samba Cluster - Opérationnel

Globus Data Transfer - Opérationnel

Historique des incidents

juin 2023

Mounting storage from campus network
  • Résolu
    UTC
    Résolu

    This incident has been resolved.

  • Détecté
    UTC
    Détecté

    Due to a HUIT networking issue, mounting storage from wired campus connections may not work.

    We are working with HUIT to resolve this.

    In the meantime, you should still be able to mount storage while connected to the RC VPN.

Emergency maintenance 6/20/23 - Cannon Slurm Scheduler
  • Pas encore commencé
    juin 20, 2023 à 11:00UTC
    Pas encore commencé
    juin 20, 2023 à 11:00UTC

    WHAT
    Emergency Cannon cluster scheduler maintenance Tuesday June 20th from 7am - 11am (actual end time estimated). Running jobs will be paused during maintenance.

    Note: FASSE is unaffected by this.

    WHY
    As most of you know, there are on-going issues with the Cannon scheduler after the downtime and OS upgrades. This behavior did not show up in our earlier testing and only manifested once we were under load from user jobs again.

    After investigation we found that the scheduler is oscillating between two states every hour. During even hours the scheduler would enter a high thread state where all the traffic would pile up, causing any queries to either take a long time or time out. During odd hours the scheduler would enter a low thread state and operate normally. During periods where the scheduler is responding, jobs can be submitted and should then run normally.

    Unfortunately the root cause is still not known, and we are working closely with the vendor, SchedMD, to find a solution to return the cluster to stability. The version of slurm that we are running is 22.05.7, which while very stable — we have been running on it for the past 5 months with no problems — appears to be having issues at scale on Rocky 8.

    There is a newer version of the scheduler available, 23.02.3. This new version has a host of improvements. While the specific issue described above is not listed as one of the items fixed in this version, we anticipate that the upgrade may fix this issue owing to the various code improvements contained therein.

    Barring any solutions found over the long weekend, we intend to hold an emergency maintenance period on Tuesday June 20th starting at 7am. Running jobs will be paused. We will return Cannon to service as soon as we've completed work. Status of this maintenance will be tracked here on our status page.

    We thank you for your patience and understanding. Our team is hard at work answering your tickets and making sure that the cluster is back to full health. Also as a reminder that we hold additional Office Hours each week in June. See: https://www.rc.fas.harvard.edu/training/office-hours/

    FAS Research Computing
    https://www.rc.fas.harvard.edu/
    https://docs.rc.fas.harvard.edu/
    https://status.rc.fas.harvard.edu/

  • Terminé
    juin 20, 2023 à 01:16UTC
    Terminé
    juin 20, 2023 à 01:16UTC

    CANCELLED

    Emergency maintenance has has been cancelled. A solution to the scheduler issue has been found and Slurm is once again operating normally.

    We thank you for your patience and understanding.

Slurm scheduler slowness/delays
  • Résolu
    UTC
    Résolu

    This incident has been resolved.

    A solution to the Slurm scheduler issue has been found and implemented. The scheduler is now operating normally.

    Thank you for your patience and understanding.

  • Mettre à jour
    UTC
    Mettre à jour

    PLEASE SEE EMERGENCY MAINTENANCE NOTICE BELOW
    Emergency maintenance Tuesday June 20th 7am-11am. Running jobs will be paused.

    We are still working on finding the root cause of the scheduler slowness and timeouts. We are working with SchedMD and have sent diagnostic information for analysis.

    At this time the scheduler is up but may become unresponsive to your commands at times. You can wait and retry them.

    Jobs, once launched, should run as expected.

    OOD/VDI are working, but job submission may also be affected.

    If you receive salloc: error: Job submit/allocate failed: Invalid account or account/partition combination specified, waiting a moment and re-submitting your job again should result in a successful submission.

    Addendum: It was determined that the issue is cyclical and you will have the best luck during odd hours (e.g. - after 1pm, before 2pm) and less success interfacing with the scheduler during even hours (e.g. - after 2pm, before 3pm).

    We regret the impact this is having on your work. Updates will follow as we have them.

  • Mettre à jour
    UTC
    Mettre à jour

    We are still working on finding the root cause of the slowness and are working with the vendor to troubleshoot.

    At this time the scheduler is up but may become responsive to your commands at times. Jobs, once launched, should run as expected.

  • Mettre à jour
    UTC
    Mettre à jour

    Slurm is still being overwhelmed with large numbers of requests. VDI timeouts/502 errors, interactive job connections, job submissions, are all impacted.

    We are in contact with the Slurm developers to find a better resolution.

  • Surveillé
    UTC
    Surveillé

    We have implemented updates and the scheduler appears stable and responsive.
    We will continue to monitor and update this incident.

  • Détecté
    UTC
    Détecté

    We are currently investigating performance issues with the scheduler. This manifests in slowness when submitting jobs, job submission fails some percentage of the time, and timeout errors (including OOD/VDI).

    Also, some jobs may not exit cleanly putting nodes into drain status.

    We are working to resolve this issue. Updates to follow.

Cannon and FASSE OOD (VDI) issues
  • Résolu
    UTC
    Résolu

    This incident has been resolved. FASSE and Cannon OOD are both available.

  • Surveillé
    UTC
    Surveillé

    Cannon and FASSE OOD are not fully operational due to slurm being slow. With slurm slow, the apps wait too long to be launched and it can throw 502 errors

  • Identifié
    UTC
    Identifié

    Cannon OOD is operational.

    FASSE OOD is still experiencing issues and showing errors for new connections. Existing connections are working.
    WIP

  • Détecté
    UTC
    Détecté

    We are currently investigating issues with OOD (OpenOnDemand aka VDI).

fasselogin nodes
  • Résolu
    UTC
    Résolu

    Login to fasselogin is working. Additional login nodes are not in the rotation and will be added back once they are healthy again. But normal login is working.

  • Détecté
    UTC
    Détecté

    We are currently investigating an issue with fasse login nodes (fasselogin) timing out or not connectring.

mai 2023

Aucun incident signalé ce mois-ci

avr. 2023

MGHPCC - issues affecting some connectivity
  • Résolu
    UTC
    Résolu

    This incident has been resolved.

    Network admins found that one of the line cards on the Nexus core switch was not functioning correctly and dropping IPv4 traffic randomly. The card will be replaced as soon as possible. In the meantime, it has been isolated and connectivity has returned to normal

    Thanks for your patience.
    FASRC

  • Détecté
    UTC
    Détecté

    We are currently investigating this incident. The majority of services appear to be working normally, but reports of random storage or service connectivity indicate an issue with networking or power at MGHPCC.

    Updates to follow as we know more.

Harvard DNS continues to experience outages
  • Résolu
    UTC
    Résolu

    As far as we are aware, Harvard's DNS issues are resolved.

  • Identifié
    UTC
    Identifié

    Harvard's DNS (name resolution system which translates hostnames to IP addresses) continues to experience degradation and periodic outages. This issue may affect access to any hostname/address ending in harvard.edu.

    While this should not affect jobs inside the cluster, it will impact other resources users are trying to access including reaching login nodes, VPN, websites, drive mapping, this status page, etc.

    This is a university-wide issue and outside FASRC's control. We will continue to monitor the situation, but for mire information please see Harvard's status page (which may also be inaccessible at times): https://status.huit.harvard.edu/

FASRC Monthly Maintenance April 3rd 2023 7am-11am
  • Terminé
    avril 03, 2023 à 15:00UTC
    Terminé
    avril 03, 2023 à 15:00UTC

    Maintenance has completed successfully

  • En cours
    avril 03, 2023 à 11:00UTC
    En cours
    avril 03, 2023 à 11:00UTC

    Maintenance is now in progress

  • Pas encore commencé
    avril 03, 2023 à 11:00UTC
    Pas encore commencé
    avril 03, 2023 à 11:00UTC

    FASRC Monthly Maintenance will occur April 3rd 2023 7am-11am
    Please note that because of the June downtime, we will not hold monthly maintenance in May and July.

    JUNE DOWNTIME AND OS/SOFTWARE CHANGES
    The MGHPCC yearly power downtime will run June 5th-8th with return to full service the morning of Friday June 9th. For timeline and initially details on the downtime, our switch Rocky Linux, and changes to how we do software, please see: https://www.rc.fas.harvard.edu/blog/2023-downtime/

    Also see our Rocky 8 Transition Guide for information on the OS and upcoming test cluster. We will also offer training and consulting in the coming months (TBA). https://docs.rc.fas.harvard.edu/kb/rocky-8-transition-guide/

    Notices about the downtime with links to training, docs, and FAQs will commence in a few weeks.

    TRAINING CALENDAR
    Upcoming new user training and other sessions can be found on our training calendar: https://www.rc.fas.harvard.edu/upcoming-training/

    APRIL 3rd MAINTENANCE

    • Login node and VDI reboots
      -- Audience: Anyone logged into a a login node or VDI/OOD node
      -- Impact: Login and VDI/OOD nodes will rebooted during this maintenance window

    • RCSMB server network changes
      -- Audience: Mounted Samba shares connected at "rcsmb.rc.fas.harvard.edu"
      -- Impact: None expected. The possibility of brief disconnects
      is possible but unlikely.

    • Squid internet proxy update
      -- Audience: Any users/session/process using the internet proxy (https://docs.rc.fas.harvard.edu/kb/proxy-settings/)
      -- Impact: The proxy will be unavailable while updating

    • Scratch cleanup ( https://docs.rc.fas.harvard.edu/kb/policy-scratch/ )
      -- Audience: Cluster users
      -- Impact: Files older than 90 days will be removed.
      planning for a scheduled maintenance during that time.

    Thanks,
    FAS Research Computing
    Department and Service Catalog: https://www.rc.fas.harvard.edu/
    Documentation: https://docs.rc.fas.harvard.edu/
    Status Page: https://status.rc.fas.harvard.edu/

avr. 2023 à juin 2023

Suivant

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