Previously we have experienced several outages of our GPFS filesystem, which primarily serves the Grace cluster but which is also mounted to other clusters as well. We understand these outages have been particularly disruptive for some and we apologize for any inconvenience this has caused.
To remedy the issue, the HPC team addressed networking issues during a recent Grace outage. The GPFS filesystem currently appears stable and is currently exported to Louise and BDN. The plan is to have GPFS also mounted onto Omega but additional testing will be needed. A timeline has yet to be defined. In the meantime, if you notice any continued issues, please report to hpc@yale.edu with the time and a description of the issue.
As always, if you have any questions, concerns or comments please contact us at hpc@yale.edu
Aug 1, 2015:
The HPC team recently repaired the network configurations that had caused several unexpected GPFS file system outages. The GPFS file system resides on Grace, but is also currently mounted on Louise and BulldogN. The plan is to have GPFS mounted on Omega, but additional testing is needed.