This weekend I finished the basic setup of my new homelab! Using the 4 C6100 nodes I made a 3-way hyper-redundant Ceph cluster. This means of my 24TB Raw storage I now have 8TB of hyper-redundant storage. All drives are SATA-III for fastest speeds possible. If 1 of the drives fail the data is safe, if a node fails the data is fine and even if 2 nodes fail then the data is still there and when I swap any dead drives it will build itself back together. There are also a few new mappings for some systems on my network:

  • Raspberry Pi 5: 192.168.10.10
  • Desktop: 192.168.10.20
  • Laptop: 192.168.10.30

On Proxmox I made a 6-node cluster with the naming convention for each server being pve1, pve2 and so on this is the mapping for each node:

  • Proliant 1: pve1 (192.168.20.10)
  • Proliant 2: pve2 (192.168.20.20)
  • C6100 Node 1: pve3 (192.168.20.30)
  • C6100 Node 2: pve4 (192.168.20.40)
  • C6100 Node 3: pve5 (192.168.20.50)
  • C6100 Node 4: pve6 (192.168.20.60)

pve1 and pve2 are the main nodes for controlling the ceph cluster and the overall proxmox cluster. With a cluster I have a near limitless amount of computational power for my needs, this is the hardware specifications for each node:

  • pve1: CPU - 2 x Xeon E5-2609 v2 (8 cores), RAM - 256GB DDR3, Storage - 120GB SSD
  • pve2: CPU - 2 x Xeon E5-2609 v2 (8 cores), RAM - 256GB DDR3, Storage - 120GB SSD
  • pve3: CPU - 2 x Xeon X5660 (24 cores), RAM - 24GB DDR3, Storage - 6TB HDD
  • pve4: CPU - 2 x Xeon X5660 (24 cores), RAM - 24GB DDR3, Storage - 6TB HDD
  • pve5: CPU - 2 x Xeon X5660 (24 cores), RAM - 24GB DDR3, Storage - 6TB HDD
  • pve6: CPU - 2 x Xeon X5600 (24 cores), RAM - 24GB DDR3, Storage - 6TB HDD

The plan for the following week is to setup a few game servers for my friends.