High Performance Computing Center North
We will perform some maintenance on the /pfs/nobackup filesystem on 2013-07-08
All batch systems will be stopped and no jobs running.
Access from the login nodes will also be unavailable.
Maintenance will start 08:30 and is expected to be finished before 11:30
Fri, 2013-06-28 13:07 | Åke Sandgren
Two of the file servers for the /pfs/nobackup file system crashed around 11:00 today (May 14th). Unfortunately this resulted in a total loss of access to the file system.
We are currently restarting them again and will then slowly restart /pfs/nobackup on all clusters again.
Tue, 2013-05-14 11:17 | Åke Sandgren
Due to human error the /pfs/nobackup filesystem on Abisko had severe problems around 06:00 May 14th.
Lots of nodes lost the connection to the filesystem and jobs where killed.
We apologize for this and are working to get it back online again.
Tue, 2013-05-14 08:55 | Åke Sandgren
A new version with some bugfixes and new features for our batch system, slurm, has been released.
The new version has improvments in how the jobs are scheduled that will enhance utilization of the cluster and we will also change some other configurations for better throughput of jobs.
The upgrade is a major version and will require all jobs to be stopped during the upgrade.
The upgrade should be complete within an hour including testing of the new version on abisko.
System upgrade finished at 10:20.