Réunion pour les discussion sur le projet Plateforme de Calcul CPPM
La réunion aura lieu dans la salle Sugiton
Meeting of the 1st July, 10am
Present: Andrei, Edith, Thierry, Carlos
The Condor batch system is upgraded. ARC5 to ARC6 is ongoing: The CE marcce1 is still at ARC5 while marcce2 is already at ARC6. While the Condor uupgrade, the WNs were also upgraded with the latest OS patches
Tests of xcache at CPPM are foreseen. A file server at CPPM is mad available for the testbench pool of SEs of the ESCAPE demonstrator
ImXgam - Mathieu (access using python API)
LHCb - 3 group servers created with a shared NFS mounted disk (LHCb share). A contextualisation script is created which bootstraps the VMs with the LHCb environment (CVMFS, etc) and defines access to the group users with permanent home directories in the LHCb share disk (3 TB). More disk space can be allocated to LHCb to motivate the group migration towards the cloud environment for the group servers. This is to be used as a model for other groups.
The v100 server is intensively used by the LHCb group. The 2080 Ti server is fully functional and available for the use by the groups Matiere Noir, LHCb, ATLAS and others. Julien installed Anaconda on th GPU02 machine to ease setting the python environement by the users The last GPU card is still being delivered, Adrian is following it.
The hardware has still one year more garantie. The cluster is used with direct access to the nodes, no batch system used. Users are from CPPM but also LAM, CPT, LUPM, IRAP.
Next meeting will take place on the 16th July at 14pm