I would like to extend my wireless into my backyard. There is solar charged electric supply with which I can energize a OpenWrt router. A few years ago, I had used relayd and I found that recipe still listed here.
Is this still the preferred method? Are there any other alternatives? Please advise.
The preferred method is always https://openwrt.org/docs/guide-user/network/wifi/wifiextenders/wds, if supported on both ends - and if cables (for outdoor, fibre - eases a lot of issues, from lightning strikes to earth potential differences compared to copper) is really not an option.
I have two vlans (Lan and IOT). Can I create a trunk interface(TRNK) with bridging eth0.lan & eth0.iot and add firewall rules for Lan & IOT. And then have an wifi bridged to this interface. On the other side, the OpenWrt client device could split then service both lan and iot devices on the split up interfaces.
Is this in the possibilities? I ask because wifi is tricky to debug and figure out.
WLAN has no concept of VLANs, nor does real hardware support more than a single STA interface per radio. So if you need a wireless backhaul, the answer is "no", unless you enlist the help of tunneling mechanisms on top (e.g. GRE, B.A.T.M.A.N. or simular).
If you can use a wired backhaul, "yes, probably" depending on what's inbetween.
Be careful if WED or possibly other wifi to ethernet software bridges are running on the WDS AP or stations, for me it caused strange network problems which were solved only after disabling WED everywhere.
In my experience, with a Cudy WR3000 v1 as dumb access point and a Cudy TR3000 as WDS client (all running openwrt 24.10), hosts wired to the WDS client would successfully get DHCP, but network performance was bad and network access randomly stopped working. In the end I disabled WED and wifi bridge works great (without relayd, or bridger or anything else). I tried changing the station with other routers, such as Netgear R7800, but no luck, WDS + WED was always a bad combination.
I learned a new term WED (Wireless Ethernet Dispatch) after I searched for it!! My use case is not a lot of traffic so I am not hoping to add this layer of complexity. However, I will now study more and learn about this beyond this encounter! Thanks again.
Fascinating. I had no idea that Wireless did not have tagged and untagged packets. I suspect this is the reason you are suggesting encapsulation and then have the interface outside of wireless domain do the discovery of tagged vs untagged.