Kia ora team,
I’m looking for a little advice about how best to setup parts of an XCP-ng install. Its just in a homelab for a bit of fun and to learn on, I’ve managed to follow Tom’s videos and get both XCP-ng running on an old Dell R730, and XO running on an old Qotom box. My (simplified) network layout is:
- Vlan 1: Default - nothing on it
- Vlan 2: Management - used for access to administrative interfaces and network control plane
- Vlan 3: Trusted network - intend to deploy VM’s to addresses on this network once they are working the way I want
- Vlan 4: Lab - locked down, can get to the internet but not to any other VLAN (addresses can reach each other). VM’s start here until they are working as intended then move to Vlan 3.
- Vlan 5: Storage - not routed on pfSense, just configured within XCP-ng and on the UniFi switches. Storage target is a Synology NAS.
What I’d appreciate advice on is:
- The server has four ports; 2x10G Ethernet, and 2x 1G Ethernet. What are the relative pro’s and con’s of keeping the physical interfaces independent of one another (i.e. one network/Vlan per PIF) rather than LAGGing the 2x10G and then having everything available over that connection.
- I understand the concept of not routing storage, but the practicalities of setting it up haven’t been so clear. The storage is iSCSI on a Synology NAS. Was easy enough to setup the iSCSI but how does XCP-ng know to access it via the correct PIF? Do I need to setup a static route, or have I missed somewhere to configure the path to the storage?
Thanks in advance.
Andy