> On 16 Feb 2020, at 12:58, Neal Becker wrote: > > I'm testing wireguard > wireguard-0.0.20191219-2.fc31.x86_64 > between a Fedora 31 client and server, comparing to openvpn. > > Openvpn is running between a linux client outside my lan and a server on my router, which is running dd-wrt. > I'm pinging between that linux client and another linux client within my lan. > > wireguard is running between the same linux client outside my lan and the same other linux client within my lan. This > time router is simply forwarding packets via NAT. > > Openvpn ping times are much lower (about 10ms) and much lower variance than wireguard. Wireguard pings > are all over the place. > > Packets coming in from the WAN are traversing some firewall that I don't control, which may be affecting results. Openvpn > is config to use udp. > > Any ideas? I can think of two ways to investigate this. Use tcpdump on the router to see the timing of the packets on the wg0 device. How quickly do you see the ICMP echo get answered? Use traceroute from the outside to find the ping times of all the hops on the path. Is there a significant difference between the ping time on the edge of the router and the wg0 device ping time? Barry > > Thanks, > Neal > > -- > Those who don't understand recursion are doomed to repeat it > _______________________________________________ > WireGuard mailing list > WireGuard@lists.zx2c4.com > https://lists.zx2c4.com/mailman/listinfo/wireguard