Scholar Cluster Maintenance
June 15, 2020 8:00am – June 19, 2020 2:35pm
As of 2:35pm, engineers have completed all maintenance tasks and have returned the Scholar cluster back to normal service. All queues have been enabled and jobs have resumed scheduling.
As a result of this maintenance you will see several important changes outlined below.
- The hardware that provides home directories storage for the Scholar cluster has been upgraded to provide better performance.
- These new Scholar home directories are dedicated to Scholar cluster only, and are completely separate and independent from the main Research Computing home directories on other systems and Community Clusters.
- There will be no automatic data migration between main Research Computing home directories and new Scholar ones. However, to avoid disruption to current classes, we have performed the migration for all affected Summer-2020 Scholar users.
- Users who have accounts on other clusters and wish to fully or partially copy their Research Computing home directories into their Scholar ones, can do so at their discretion using provided instructions.
Other noteworthy changes:
- New versions of multiple software packages have been deployed, including Thinlinc, Slurm, R, OpenMPI.
- A seven-day timeout has been implemented for idle Thinlinc sessions.
As always, thank you for using Scholar and your flexibility during this maintenance. Please report any issues to firstname.lastname@example.org
UPDATE: June 15, 2020 8:55am
The Scholar cluster is undergoing regular inter-semester maintenance and upgrades this week (June 15 – 19, 2020). All jobs which cannot complete before the start will be held queued during this time, and no one will be able to log in or use Scholar for interactive work during this maintenance.
One of the biggest changes is that we will upgrade the hardware that provides home storage for the Scholar cluster. When the process is completed, each Scholar user will get a new home directory of better performance.
This home directory will be completely separate from the main RCAC home directory. After the maintenance, users will be able to copy their main RCAC home contents to Scholar using the provided
copy-rcac-home tool on Scholar front-end nodes. In addition, selective file transfer will be available using Globus or scp/sftp/rsync from other clusters or data.rcac.purdue.edu. The detailed steps will be outlined in the Scholar User Guide.
New versions of multiple software packages will be deployed, including ThinLinc, Slurm, R, OpenMPI. Additionally, a seven-day timeout will be implemented for idle ThinLinc sessions.
As always, please contact us at email@example.com for any questions, problems, or concerns.
ORIGINAL: May 27, 2020 11:28am
The Scholar cluster will be taken down for regular inter-semester maintenance and upgrades starting at Monday, June 15th, 2020 at 8:00am. All jobs which cannot complete before then will be held queued during this time, and no one will be able to log in or use Scholar for interactive work during this maintenance.
Scholar is planned to return to regular service by Friday, June 19th, 2020 at 2:35pm.