Comment # 5 on bug 862739 from Per Jessen

It's been about 17 months since I opened this report, I guess we switched using
p2p although I wrote "mode subnet" above. This is the current server config:

local minerva.enidan.com
dev tun
topology p2p
server 192.168.13.0 255.255.255.0
push "route 192.168.0.0 255.255.240.0"
user nobody
group nobody
daemon
persist-key
persist-tun
client-config-dir /etc/openvpn/clients
ccd-exclusive
keepalive 10 120
ca ca.crt
cert server.crt
key server.key
dh dh1024.pem

A typical client config from the server:

ifconfig-push 192.168.13.9 192.168.13.10 255.255.255.0

Local client config:

client
dev tun
proto udp
remote minerva.enidan.com 1194
resolv-retry infinite
nobind
user nobody
group nobody
persist-key
persist-tun
ca ca.crt
cert office74.local.net.crt
key office74.local.net.key
ns-cert-type server
verb 3
script-security 2

When I start a client, I get this interface config:

4: tun0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
state UNKNOWN qlen 100
    link/none 
    inet 255.255.255.0 peer 192.168.13.10/32 scope global tun0
       valid_lft forever preferred_lft forever

The route config looks okay:

192.168.0.0/20 via 192.168.13.10 dev tun0
192.168.13.10 dev tun0  proto kernel  scope link  src 255.255.255.0

Client openvpn version: openvpn-2.2.2-9.9.1.x86_64
server openvpn version: openvpn-2.3.4-2.4.1.i586

I'll attach the client log in a minute, but looking at the interface getting
configured, it's clearly bonkers:

/bin/ip addr add dev tun0 local 255.255.255.0 peer 192.168.13.10