I have a simple home lab setup with 3 nodes, all connected to a 2.5G switch, the latency between all nodes is around 0.45 ms, the overall workload of the LAN is couple of iPhone(s), iPad(s) and 15 IoT (mostly ring cameras, thermostat, Roomba, TV) and the Proxmox Cluster with Ceph. All devices but the Proxmox nodes and the wireless AP, are all 1G or less, all IoT(s) are even lower speed. Traffic is normal to low, most of the day.
I have limited availability of switches and ports on the nodes, that's the reason for this "all-in" setup in one network. I might be able to figure out a way to create a separate network for the Cluster, Migration (only perform migration when I have to reboot a node, usually with Proxmox updates) and Ceph (only 2 VMs are using Ceph disks, OpenWrt and Adguard, they run most of their processes in RAM, therefore low disk activity).
Having all this information, you understand why I did it the way I did it, it is working flawlessly, I have never seen any issue in the Cluster or Ceph dashboards or while doing a migration, does it make sense to separate them? Thanks
I have limited availability of switches and ports on the nodes, that's the reason for this "all-in" setup in one network. I might be able to figure out a way to create a separate network for the Cluster, Migration (only perform migration when I have to reboot a node, usually with Proxmox updates) and Ceph (only 2 VMs are using Ceph disks, OpenWrt and Adguard, they run most of their processes in RAM, therefore low disk activity).
Having all this information, you understand why I did it the way I did it, it is working flawlessly, I have never seen any issue in the Cluster or Ceph dashboards or while doing a migration, does it make sense to separate them? Thanks