CSCS Scheduled Maintenance – Wednesday April 10, 2013

Today, April 10th, is the regular scheduled maintenance day for CSCS systems. The following systems/services will be affected, with the expected downtimes as indicated:

* Network : short interruptions of the Infiniband network (impacting  /home, /apps/, /store and /project) might be experienced from 10:30 AM  to Noon.

* Cooling : due to an intervention on the cooling infrastructure during that day, all CRAY systems (with the exception of Albis/Lema) will be  affected from 8:00AM to 5:00PM.

* Slurm: upgrade from version 2.3.4 to version 2.5.4 across the entire centre.

* Front-end system, Ela: 08:00AM – Noon for slurm, scratch filesystems and PrgEnv  interventions. BE AWARE: As announced already on March 5th and subsequently on March  25th, we will introduce new external login nodes for Monte Rosa. As a consequence, the role of Ela will change: therefore the CRAY programming environments and Slurm will be removed from the front-end. Also, the /scratch of Monte Rosa, Todi, Rothorn, Eiger, Julier and Pilatus will be unmounted from Ela. As a replacement, the new Monte Rosa (ssh and Piz Daint (ssh login nodes can be used for  compilation and job submission. All your compilations and batch scripts should run as before.

* Cray XC30, Piz Daint external login nodes: 08:00AM – Noon slurm upgrade.

* Cray XC30, Piz Daint: 08:00AM – 8:00PM cooling testing, slurm upgrade and software  patches.

* Cray XE6/XK7/XMT, Monte Rosa/Todi/Matterhorn: 08:00AM – 5:00PM cooling testing and slurm upgrade.

* Intel SB Cluster, Pilatus: 8:00AM – 4:00PM kernel security patch and power supply interventions.

* Visualization and Post-processing, Eiger/Piz Julier : 08:00AM – Noon slurm upgrade.

* iDataPlex, Castor: 08:00AM – Noon PrgEnv upgrade and low level software interventions.

* SGI Altix UV, Rothorn: 08:00AM – Noon  slurm upgrade.


Leave a Reply