System Downtime October 8, 2019

Date: 
Tuesday, October 8, 2019 - 7:00am to 5:00pm
Location: 

Ohio Supercomputer Center

Updated at 4:50 PM Oct 8, 2019: 

Downtime for all HPC systems has been extended to 7 p.m., Tuesday, October 8, 2019

Original Post:

A downtime for all HPC systems is scheduled from 7 a.m. to 5 p.m., Tuesday, October 8, 2019. The downtime will affect the Pitzer, Ruby and Owens Clusters, web portals, and HPC file servers. Login services, including my.osc.edu, access to storage, and license server for state-wide license software hosted by OSC will not be available during this time.

In preparation for the downtime, the batch scheduler will begin holding jobs that cannot be completed before 7 a.m., October 8, 2019. Jobs that are not started will be held until after the downtime and then started once the system is returned to production status.
 
Highlights of the downtime activities include:

  • Upgrades to OSC core Ethernet networks
  • Preventative maintenance and upgrades for the storage environment
  • Preventative maintenance and upgrades for the InfiniBand fabric
  • Deployment of a new GPU reservation scheme on Owens and Pitzer HPC clusters
  • OS updates for the Owens, Pitzer, and Ruby HPC clusters
  • Installation of Cuda 10.1.168 on Pitzer, Owens and Ruby HPC clusters
  • Removal of BWA versions 0.1.17 and 0.17.13 from Owens and Pitzer due to a software vulnerability (for more details, see this link)
  • Removal of OpenMPI versions 1.10-hpcx and 2.0-hpcx from Owens (these versions have errors and no longer work correctly in our environment)
  • Replacement of Q-chem 5.1.1 by Q-chem 5.2.1 as the default version on Ruby, Owens and Pitzer HPC clusters (Note: all Qchem 5.1 variants will become unavailable on Dec 1st, 2019 due to a change in Q-chem license management)

To stay up to date on system notices, follow @HPCNotices  on Twitter. As always, you can contact us at OSC Help