This repository will contain automated scripts for settings up our training or production infrastructure at least in the cloud (we'll add bare-metal ASAP).
- automated setup scripts for new clean machines
- running kubernetes clusters
- running gpu boosted kubernetes clusters
- reproducible infrastructure throught terraform
- Nvidia CUDA & CUDNN setup
- Connecting & mapping external services
- VM initialization scripts
- DNS mapping (for route53)
- Terraform configurations
- Static IP addresses allocation
- automated self destruction ?????
- setting up physical networks
- settings up physical volumes & storage environnement
- bare metal security aspects