On Saturday 23rd of September, we will be performing essential maintenance on the file servers that host QRIScloud RDS Tier2 collections. (This will not affect Tier3 or GPFS collections.)
We anticipate that the maintenance will require a service outage of up to 4 hours. During this time, data in Tier2 collections will be inaccessible, and any systems that attempt to read or write to the collections will be blocked:
- In most cases, a command, job or file transfer that attempts to access a file will freeze, and then resume automatically once the outage is over.
- For HPC users, an up-to 4 hour "freeze" may cause running jobs to exceed their walltime limits and fail. If this is a concern, you can use the "qhold" and "qrls" commands to place a manual hold on your jobs and then release them after the outage.
Following the server outage, it is possible that some systems will experience stale NFS mounts:
- QRIScloud operations will deal with these problems for the Standard Access services, and on the HPC systems.
- Users with collections that are configured for NFS Only access, may need to manually unmount and remount the collection to clear the problem.
If you observe problems during or following the outage, please report them to QRIScloud Support via a support ticket.
We apologize for the any disruption caused by this outage.
UPDATE - 2017-09-23 - 12:30 - The server outage been completed, and the standard access services have all recovered.