I need some help.
I have a stange wireguard issue.
I subnetted out a /24 into multiple /29 locally. (Announced as a /24 over bird2 to other bgp peers, only subnetted locally on server).
Locally mounted IPs have no problems. Every /29 that is locally mounted doesn't have issues to ping eachother, the outside world, traceroute shows to the bgp upstream, and direct to the IP. Looks good.Example of traceroute: outside client -> bgp peer upstream -> IP.
However wireguard peers are with issues. They can't talk to the locally mounted IPs even they are in different /29s. And they can't even ping to their gateway on the wireguard server. but they ping and traceroute to outside IPs no issues.Example traceroute from wireguard = client ip -> wireguard mounted IP in the wireguared config with /29 they are in -> bgp peer upstream -> etc -> destinationExample traceroute from outside to wireguard client = outside client -> bgp peer upstream -> server outside IP -> client IP (notice in the wireguard case, the main IP of the bgp server shows, which doesn't for locally mounted IPs)
If the server end, directly on SSH pings a wireguard client, it replies and no issues.If a wireguard client pings to ANY IP in the parent /24, no reply, and traceroute shows it goes to upstream bgp and back and forth and loops.
Any info on this?
Note, wireguard and locally mounted IPs are externally routed IPs, direcly routed over wiregaurd.
