proxmox:
version: pve-manager/7.2-3/c743d6c1
router:
kernel: Linux 4.4.241 #1327 SMP Tue Nov 3 09:06:16 +03 2020 armv7l
build: DD-WRT v3.0-r44715 std (11/03/20)
hardware: Asus RT-AC5300
Howdy, folks. I'm seeing some strange behavior from the DHCP server servicing my embedded switch. I have a Proxmox server attached to a switch port. Proxmox is configured for bridging with bridge IP assigned via DHCP. VMs mapped to that bridge have unique MAC addresses and are requesting addresses via DHCP also.
As the VMs boot the router is reassigning the same LAN IP address to each VM despite each of them having a distinct MAC. There's nothing in the logs which suggests, for instance, that the initial DHCP request is going out with the bridge MAC and thus that's the source of the re-used IP. Generally, there's nothing in the router logs nor in Proxmox logs to suggest that either is unhappy with the DHCP process though clearly the VMs end up not being happy about the situation.
I dug into STP settings and didn't see anything special that needed to be set on the router to allow for multiple MAC addresses on a single switch port.
I also didn't see anything special that might need to be set in Proxmox bridge networking to play nicely with an upstream non-VLAN-aware switch.
VLANs are enable neither on the router nor on the Proxmox server. I have the option to store MAC addresses in NVRAM turned off because it seems to prevent the WAN port on the router from getting a DHCP address from ISP for some undetermined region.
Any thoughts? At this point I'm considering just virtualizing the firewall/router as a work-around because this has me pretty baffled.
Thx
version: pve-manager/7.2-3/c743d6c1
router:
kernel: Linux 4.4.241 #1327 SMP Tue Nov 3 09:06:16 +03 2020 armv7l
build: DD-WRT v3.0-r44715 std (11/03/20)
hardware: Asus RT-AC5300
Howdy, folks. I'm seeing some strange behavior from the DHCP server servicing my embedded switch. I have a Proxmox server attached to a switch port. Proxmox is configured for bridging with bridge IP assigned via DHCP. VMs mapped to that bridge have unique MAC addresses and are requesting addresses via DHCP also.
As the VMs boot the router is reassigning the same LAN IP address to each VM despite each of them having a distinct MAC. There's nothing in the logs which suggests, for instance, that the initial DHCP request is going out with the bridge MAC and thus that's the source of the re-used IP. Generally, there's nothing in the router logs nor in Proxmox logs to suggest that either is unhappy with the DHCP process though clearly the VMs end up not being happy about the situation.
I dug into STP settings and didn't see anything special that needed to be set on the router to allow for multiple MAC addresses on a single switch port.
I also didn't see anything special that might need to be set in Proxmox bridge networking to play nicely with an upstream non-VLAN-aware switch.
VLANs are enable neither on the router nor on the Proxmox server. I have the option to store MAC addresses in NVRAM turned off because it seems to prevent the WAN port on the router from getting a DHCP address from ISP for some undetermined region.
Any thoughts? At this point I'm considering just virtualizing the firewall/router as a work-around because this has me pretty baffled.
Thx