Hi there, first post. I'm running OpenWrt 18.06.1 r7258-5eb055306f / LuCI openwrt-18.06 branch (git-18.228.31946-f64b152) on a WRT3200ACM and I'm trying to create an isolated, wired vlan (vlan 3 in my case). I've configured one switchport for that vlan thusly;
but it makes no difference. My regular lan (vlan 1) is 192.168.0.0/24, so I create a new interface, assign it to eth0.3 and give it a static address of .1 on 192.168.1.1/24. I also set up a DHCP server for this range. So far so good.
I plug a laptop into the port designated for vlan 3, and it gets an address on that network - 192.168.1.175, and the correct mask and gateway. So far so good. Except that laptop cannot ping its gateway 1.1 (Why? It should just be straight layer 2), and devices on vlan 1 can ping the interface on vlan 3 ie my PC at 192.168.0.5 can ping 192.168.1.1 as well as its own gateway at 0.1!! (Why? I have not even gotten to routes or FW zones yet?). I'm completely baffled by this. Do vlans actually work on this device's switch? If layer 2 does not work as expected there is not much point in going further. Any ideas are appreciated!
So not wanting to get ahead of myself, why can't my laptop with its DHCP assigned 192.168.1.175 ping its gateway at 192.168.1.1? It's the same subnet so no routing or FW zones should be needed. Do I need to allow ICMP responses from that interface to it's own vlan somewhere?
Whatever devices connected to that vlan are likely going to need to have the vlan ID set to 3 on the device's interface for traffic to route correctly.
Just to verify, you are plugging into LAN port 1 [sw.port 3] on the router?
Hi again. Thanks for the reply. Yes, I am plugging into LAN1. I'm connecting that to a Cisco 2970G interface configured as an access port on its vlan 3, though I have tried the laptop directly to the Linksys with the same result. I was hoping this device might do dot1q trunking but I know that was a bit much to ask. I'll be happy enough if I can get different individual tagged vlans on the switch ports of the 3200 and distribute them on the Cisco.
Hey. Thanks, I thought it should. I've tried configuring the downstream switchport as a trunk but no luck either. PVID option did not seem to help either. From my reading here https://openwrt.org/docs/guide-user/network/vlan/switch_configuration it looks like dot1q should be the default protocol on ports with multiple vlans. I've tried this configuration with a dot1q trunk on the Cisco connected to LAN1 on the Linksys, but still get the same funky behavior.
I'm recently retired from a career in enterprise networking so switching and routing is second nature but I'm used to Cisco devices. I've only had this Linksys a couple days so perhaps I need to spend more time reading TFM.
Still, it just seems completely odd that a device on vlan 1 can ping the eth0.3 gateway address, and a device on vlan 3 can't.
So basically any subinterface I create on eth0.x is pingable from vlan 1. I assigned 172.16.0.1 to eth0.4 with no bridging or anything and I can immediately ping that address from clients the in 192.168.0.0/24 subnet. Its like the CPU itself is ignoring the tagging......
I'm not sure about the other issues, but after looking at my old managed switch config, pinging vlan 3's gateway should be fixed by adding option gateway 'xxx.xxx.x.x' to dmz's interface options.