HP Z640 Workstation / Old Proxmox Node
This was the first serious Proxmox proof-of-concept machine for the homelab. It proved that the Terraform + Ansible + VM-per-service direction was worth doing before I committed to the EPYC build.
It is no longer the main compute node. The EPYC 7542 server has taken over the heavy VM, GPU, and service hosting role. The Z640 is now better thought of as a spare/secondary virtualization box, a testing machine, or a place to try migrations before I do something dumb to the main node.
Hardware
- CPU: 2x E5-2670v3
- RAM: 256 GB DDR4 RDIMM
- Storage: 512GB BX500 SSD
- Storage: 2TB BX500 SSD
- Storage: Asus Hyper M.2 pcie4.0 x4 expansion card
- Storage: 2x 4 TB 990 Evo Plus
- Storage: 2x 2 TB 990 Pro
Active Services
Most if not all of my services are spun up using terraform, and configured with ansible scripts. Once I refactor the playbooks into proper playbook formatting, testing with AWX, and doing a full lab config/restore from the scripts. With the end goal of running the configuration on a purpose built server and provisioining that to achieve full functionality.Raw VMs
- Recursive DNS combo - Adguard+Unbound
- Authoratative DNS - BIND9
- Internal Certs / reverse proxy - Caddy
- Automation - Jenkins
- Automation - Rancher (to manage k3s, and run AWX and its dependencies)
- Monitoring - Prometheus / Grafana
- Git - Gitea
- Terraform worker node
- Terraform worker node