Requirements for deployment

• A kubernetes cluster, API > v1.23
  • CSI with ReadWriteMany persistent volumes
  • Ingress controler and certificate manager

Typical minimal base platform: frontend and dispatcher + Contributed notebook backends

• 50 CPU, 100 Gb RAM, 200Gb persistent storage for container images

HPC cluster link

• HPC cluster, e.g. slurm scheduler, and service account (possibly serveral)
• typical capacity 50 jobs to run at the same time
• egress from the HPC cluster to kubernetes cluster
• shared storage for intermediate results, 1 Tb

For breakdown and the case of INTEGRAL analysis, see there.