Update: This upgrade is taking substantially longer than expected. Currently we expect the system be available again beginning on August 8, 2016.
At approximately 11:50 this morning service was interrupted at the ITF data hall. We presume a power failure, but this has not been confirmed just yet. Jobs that were running on Neon have stopped as a result.
HPC staff are bringing the cluster back online at this time.
The /nfsscratch filesystem on Neon is nearly full and the snapshot from initial cleaning (prior to July 1) will be deleted on Friday July 15 to reclaim free space.
Effective July 1, 2016, the 30-day file retention policy for networked scratch filesystems on both Helium and Neon will change.
The FhGFS (/scratch) file system has been unmounted on compute nodes as discussed in the entry, Helium FhGFS scratch decommissioning | High Performance Computing.