Hi folks,
a next question about Linux bridges:
I'll have a NAS virtualized on the Proxmox host. On that VM I'll create an NFS share, which will be used by the Proxmox host.
The network is 10.10.0.0/24, all servers (VM and physical) are in that network and physical ressources are connected via a Ubiquiti switch.
That means vmbr0 (and thus the proxmox host) has IP 10.10.0.100. The NAS VM has 10.10.0.20. If the host now accesses data on that NAS via NFS, will that traffic go out to the physical switch and therefore be limited to 1 GBit/s uplink speed, or will it stay inside the virtual network an can potentially use higher speeds?
If it goes out to the pysical switch, does ist make sense to create a second bridge (vmbr1) and assign this bridge an IP of a different subnet?
Like vmbr1 has IP 10.10.1.1/30 with no slave ports and the NAS becoms a second virtual NIC with IP 10.10.1.2/30. Would the traffic in this case stay virtual?
Would it nevertheless make sense to separate VM-host traffic anyway?
Thx!
a next question about Linux bridges:
I'll have a NAS virtualized on the Proxmox host. On that VM I'll create an NFS share, which will be used by the Proxmox host.
The network is 10.10.0.0/24, all servers (VM and physical) are in that network and physical ressources are connected via a Ubiquiti switch.
That means vmbr0 (and thus the proxmox host) has IP 10.10.0.100. The NAS VM has 10.10.0.20. If the host now accesses data on that NAS via NFS, will that traffic go out to the physical switch and therefore be limited to 1 GBit/s uplink speed, or will it stay inside the virtual network an can potentially use higher speeds?
If it goes out to the pysical switch, does ist make sense to create a second bridge (vmbr1) and assign this bridge an IP of a different subnet?
Like vmbr1 has IP 10.10.1.1/30 with no slave ports and the NAS becoms a second virtual NIC with IP 10.10.1.2/30. Would the traffic in this case stay virtual?
Would it nevertheless make sense to separate VM-host traffic anyway?
Thx!
Last edited: