Two instances, both running CentOS 6 and strongSwan 5.4.0. The kernel is
2.6.32-754.17.1.el6.x86_64
Both instances are behind NAT (AWS, separate VPCs, different EIPs
assigned to each instance - the two private IP ranges are completely
separate). This is ipsec.conf on one instance (the other instance's
config is similar, just mirrored left/right):
###
config setup
strictcrlpolicy=yes
uniqueids = yes
conn %default
authby = psk
auto = start
compress = yes
dpdaction = restart
dpddelay = 5s
dpdtimeout = 30s
esp = aes256-sha512-modp4096
forceencaps = yes
fragmentation = yes
ike = aes256-sha512-modp4096
keyexchange = ikev2
type = tunnel
conn site1-site2
left = %any
leftid = site1
leftsubnet = 10.0.1.254/32
right = public-IP-of-other-instance
rightid = site2
rightsubnet = 10.0.2.254/32
###
leftsubnet and rightsubnet are the private IPs assigned to eth0 on the
VPN instances.
Once the IPSec tunnel is established, I can ping eth0 of instance
site2-vpn from instance site1-vpn, and viceversa. So the IPSec tunnel
works fine. I've used strongSwan before, I'm familiar with this part.
###
[root@site1-vpn ~]# ping -n 10.0.2.254
PING 10.0.2.254 (10.0.2.254) 56(84) bytes of data.
64 bytes from 10.0.2.254: icmp_seq=1 ttl=64 time=0.606 ms
64 bytes from 10.0.2.254: icmp_seq=2 ttl=64 time=0.690 ms
64 bytes from 10.0.2.254: icmp_seq=3 ttl=64 time=0.685 ms
###
My initial goal is to create a GRE tunnel between the two instances,
then run dynamic routing (BGP, OSPF, whatever works) and let dynamic
routing take care of adding/removing routes between VPN nodes (which are
default gateways to their local networks). The ultimate goal is to have
a mesh of VPN nodes with multiple tunnels in between and dynamic
routing. I want strongSwan to handle the minimum amount of routing
needed to create the GRE tunnels, and then let dynamic routing handle
the bulk of the routing tables on top of GRE.
But first, I need to establish one GRE tunnel between two nodes.
This is how I create the GRE tunnel (the local and remote IPs are the
eth0 addresses on the VPN instances):
###
# site1-vpn
ip tunnel add gre0 local 10.0.1.254 remote 10.0.2.254 mode gre ttl 255
ip link set gre0 up
ip addr add 10.2.0.1/24 dev gre0
# site2-vpn
ip tunnel add gre0 local 10.0.2.254 remote 10.0.1.254 mode gre ttl 255
ip link set gre0 up
ip addr add 10.2.0.2/24 dev gre0
###
But when I try to ping the other end of the tunnel, it does not work:
###
[root@site1-vpn ~]# ping -n 10.2.0.2
PING 10.2.0.2 (10.2.0.2) 56(84) bytes of data.
ping: sendmsg: Invalid argument
ping: sendmsg: Invalid argument
ping: sendmsg: Invalid argument
###
tcpdump on the other instance shows no activity - no packets are even
being sent to the other side via the IPSec tunnel. But the IPSec tunnel
continues to work, I can ping eth0 IPs across the tunnel just fine.
This is networking on one instance (the other is similar, just
mirrored):
###
[root@site1-vpn ~]# ip route
10.0.1.128/25 dev eth0 proto kernel scope link src 10.0.1.254
10.2.0.0/24 dev gre0 proto kernel scope link src 10.2.0.1
default via 10.0.1.129 dev eth0
[root@site1-vpn ~]# ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: mtu 9001 qdisc pfifo_fast
state UP qlen 1000
link/ether 02:96:54:76:52:80 brd ff:ff:ff:ff:ff:ff
inet 10.0.1.254/25 brd 10.0.1.255 scope global eth0
inet6 fe80::96:54ff:fe76:5280/64 scope link
valid_lft forever preferred_lft forever
3: gre0: mtu 1476 qdisc noqueue state UNKNOWN
link/gre 0.0.0.0 brd 0.0.0.0
inet 10.2.0.1/24 scope global gre0
4: gretap0: mtu 1476 qdisc noop state DOWN qlen
1000
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
###
One more thing - when I run "ip tunnel add" I get this warning (I don't
know whether it matters or not):
###
[root@site1-vpn ~]# ip tunnel add gre0 local 10.0.1.254 remote
10.0.2.254 mode gre ttl 255
ioctl: File exists
###
--
Florin Andrei
http://florin.myip.org/
