Monthly maintenance will occur on Monday, July 6th from 7am to 11am.
NOTICES - UPDATED
IMPORTANT! - Home directories will be physically moved on July 20th. This will require the stoppage of ALL jobs and cluster logins as accounts cannot operate without a home directory. We are targeting a full day (allotting 8-10 hrs). To compensate for this interruption, we will perform any pending maintenance tasks during this window and cancel August maintenance.
More details and a list of affected services can be found at: https://www.rc.fas.harvard.edu/rsvpmaker/home-directory-move-july-2020/
- As a part of the Markley-Boston datacenter downsizing project, FASRC will be combining physical space with HMS. From June-August, we will be moving 18 racks worth of servers to a new location. The above home directory move is a required part of this.
Globus endpoints update: We now have “Harvard FASRC Holyoke” and “Harvard FASRC Boston” endpoints. Storage for each will be mounted based on data center locations. We will be retiring the #fasrc endpoint during this maintenance.
Please note that our office hours (Wednesdays 12pm -3pm) are still online.
- The annual June power downtime at MGHPCC has been postponed until October 20-21 (shutdown begins at 6PM on 10/19).
More details here: https://www.rc.fas.harvard.edu/mghpcc-yearly-shutdown
- Slurm (20.02.3) and PMIx upgrade- * Jobs will be paused *
- Audience: Users running cluster jobs of any kind
- Impact: Runnning jobs will be paused while the scheduler is upgraded
- Impact: Pending jobs will remain pending
- Impact: New jobs cannot be scheduled until complete
- Generic Globus endpoint decommission, Boston and Holyoke endpoints active
- Audience: Globus users using original endpoint
- Impact: See Globus docs for new endpoint details - https://docs.rc.fas.harvard.edu/kb/globus-file-transfer/
- New endpoint allows direct use of lab storage
- VDI/OpenOnDemand upgrade
- Audience: All VDI/OOD users
- Impact: VDI/OpenOnDemand aka OOD unavailable during maintenance
- Impact: VDI nodes will also be rebooted
- Login node firmware upgrades
- Audience: Users of cluster login nodes
- Impact: Login nodes will be unavailable for short periods while updating
- Impact: Login nodes will also be rebooted
- Login and VDI nodes will be rebooted
- Audience: Cluster users on login and VDI nodes
- Impact: Reboots will follow work on login and VDI nodes as noted above
- Scratch cleanup ( https://docs.rc.fas.harvard.edu/kb/policy-scratch/ )
- Audience: Cluster users
- Impact: Files older than 90 days will be removed.
- Reminder: Scratch 90-day file retention purging runs occur regularly not just during maintenance periods.
FAS Research Computing
Reminder: Scratch 90-day file retention purging runs occur regularly not just during maintenance periods.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Permissions beyond the scope of this license may be available at Attribution.