Monthly maintenance October 4th, 2021 7am-11am
Regular monthly maintenance will take place Monday October 4th, 2021 7-11am
NOTICES
- Update to fairshare mechanism to help break ties within lab groups. No discernable change to the majority of cluster users. This small adjustment is weighted by a users individual usage, breaking ties internal to a group. https://docs.rc.fas.harvard.edu/kb/fairshare/#Nice
- New cluster hardware comes online Sept 30th, increasing overall computational power by 25%
- new block of 18 water cooled GPU nodes, each with NVidia SXM4 (4x A100 cards)
- new block of 36 water cooled “big mem” compute nodes, each with 2x Intel 8358 processors (64 cores total) and 512 GB of RAM
- retirement of less-efficient gear including most AMD compute and K80 GPUs
- gratis fairshare has been increased to 120 for all labs
See the blog post from Scott (or Sept 13th email) for more details: https://www.rc.fas.harvard.edu/blog/2021-cluster-upgrade/ - Holyscratch cleanup will remove many retention exceptions - Owners of that data have been notified
- Training sessions available: https://www.rc.fas.harvard.edu/upcoming-training/
- New Storage Service Center launched FY22 (July 1st 2021). See Billing FAQ https://www.rc.fas.harvard.edu/policy/billing-faq/ and list of storage tiers https://www.rc.fas.harvard.edu/services/data-storage/. Individual PIs will be contacted about setting up their billing.
- If you have a 2021/2021 publication that made use of the FASRC cluster and you don't find it listed here, please do let us know: https://www.rc.fas.harvard.edu/cluster/publications/
GENERAL MAINTENANCE
-
- NCF VDI/OoD - Retire old Jupyter/Jupyterlab apps
- Audience: NCF VID/OpenOnDemand
- Impact: Removes old apps, existing/new apps not affected.
- hsphs10 fileserver will be decommissioned on October 1st
- Audience: Users of hsphs10
- Impact: Tenants have been notified, so no expected impact
- Boston and Holyoke Isilon upgrades
- Audience: Isilon (Tier1) shares in both data canters
- Impact: None expected. Upgrades should roll across redundant nodes with no downtime
- I
QSS SID beta deploymentDeferredAudience: Beta testersImpact: None
- www.rc.fas.harvard.edu and docs.rc.fas.harvard.edu upgrades
- Audience: Visitors to either website
- Impact: Both sites will be down for approximately 10 minutes each
- Login node and VDI node reboots
- Audience: Anyone logged into a a login node or VDI/OOD node
- Impact: Login and VDI/OOD nodes will be unavailable while updating and rebooting
- 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.
Thanks,
FAS Research Computing
https://www.rc.fas.harvard.edu
https://docs.rc.fas.harvard.edu
https://status.rc.fas.harvard.eduReminder: Scratch 90-day file retention purging runs occur regularly not just during maintenance periods.
- NCF VDI/OoD - Retire old Jupyter/Jupyterlab apps