The system is back to regular service

Thursday 6 April 16:00 CET (17:00 EET)

We are pleased to let you know that LUMI is back to regular service after the entire system management stack, system firmware and programming environment were upgraded for better reliability and performance. Significant parts of the system such as compilers and MPI libraries have been updated which may require you to recompile your application.

Please note:

• if you use the LUMI software stack, activate it by loading the latest available version of the LUMI module: module load LUMI/22.08
• Slurm update introduces a breaking change for hybrid MPI+OpenMP jobs. srun will no longer read in the value of –cpus-per-task (or -c) from the job script. This means you will implicitly have to specify the number of CPUs per task:
  • srun –cpus-per-task=$SLURM_CPUS_PER_TASK <your_app>

or

• • export SRUN_CPUS_PER_TASK=$SLURM_CPUS_PER_TASK followed by srun <your_app>

Don’t hesitate to contact the LUMI User Support Team at https://lumi-supercomputer.eu/user-support/need-help/ if you need any assistance or provide feedback on lacking features in the programming environment, missing software, libraries or tools, and poor performance.

Thank you for your cooperation.