History

This page shows a history of the changes that have been made to the mass storage or to the cluster over time.

2024

2024-02-09 Setup of a new global scratch

A new scratch directory has been set up. It provides a larger space than Gallia (i.e. 308 TB) and will replace it in a near future to improve security. It is now recommended to put your temporary data there (or on the nodes local storage). Its specificities and usage instructions are documented here.
2024-02-07 Upgrade of nodes local storage

New SSDs have been installed on most of the old nodes that were still using HDDs. This increases the local storage of those nodes to provide around 1.5 to 1.8 TB, and should also improve the performance of those nodes for IO intensive tasks. For more information about the specifications of the nodes, referer to this overview.
2024-01-26 Creation of a new Slurm partition

A new Slurm queue called nextflow has been created. This partition is dedicated to the submission of the master Nextflow pipelines. To accomodate this specific workload, it has no time limit, but low resources. Hence, it no pipeline process should be submitted there.

2023-12-11 Migration to the new cluster

After a few weeks of migration, the old cluster has been made unavailable. Everyone should now use the new cluster, which requires the usage of a SSH key.
2023-11-14 Publication of a new SSH connection procedure

A future upgrade of the cluster will require a SSH key. The new procedure to create and set up such a connection is described here.
2023-09-29 Publication of a Slurm submission script wizard

To simplify the submission of Slurm jobs, a new online tool has been published here. It provides a quick and easy way to write a submission script.