summaryrefslogtreecommitdiffstats
path: root/doc/nhrpd.texi
blob: 069b46495c7c3d5b0627bb9b41c0a0148baa8fe5 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
@cindex NHRP
@node NHRP
@chapter NHRP

@command{nhrpd} is a daemon to support Next Hop Routing Protocol (NHRP).
NHRP is described in RFC2332.

NHRP is used to improve the efficiency of routing computer network
traffic over Non-Broadcast, Multiple Access (NBMA) Networks. NHRP provides
an ARP-like solution that allows a system to dynamically learn the NBMA
address of the other systems that are part of that network, allowing
these systems to directly communicate without requiring traffic to use
an intermediate hop.

Cisco Dynamic Multipoint VPN (DMVPN) is based on NHRP, and
@value{PACKAGE_NAME} nhrpd implements this scenario.

@menu
* Routing Design::
* Configuring NHRP::
* Hub Functionality::
* Integration with IKE::
* NHRP Events::
* Configuration Example::
@end menu

@node Routing Design
@section Routing Design

nhrpd never handles routing of prefixes itself. You need to run some
real routing protocol (e.g. BGP) to advertise routes over the tunnels.
What nhrpd does it establishes 'shortcut routes' that optimizes the
routing protocol to avoid going through extra nodes in NBMA GRE mesh.

nhrpd does route NHRP domain addresses individually using per-host prefixes.
This is similar to Cisco FlexVPN; but in contrast to opennhrp which uses
a generic subnet route.

To create NBMA GRE tunnel you might use the following (linux terminal
commands):
@example
@group
 ip tunnel add gre1 mode gre key 42 ttl 64
 ip addr add 10.255.255.2/32 dev gre1
 ip link set gre1 up
@end group
@end example

Note that the IP-address is assigned as host prefix to gre1. nhrpd will
automatically create additional host routes pointing to gre1 when
a connection with these hosts is established.

The gre1 subnet prefix should be announced by routing protocol from the
hub nodes (e.g. BGP 'network' announce). This allows the routing protocol
to decide which is the closest hub and determine the relay hub on prefix
basis when direct tunnel is not established.

nhrpd will redistribute directly connected neighbors to zebra. Within
hub nodes, these routes should be internally redistributed using some
routing protocol (e.g. iBGP) to allow hubs to be able to relay all traffic.

This can be achieved in hubs with the following bgp configuration (network
command defines the GRE subnet):
@example
@group
router bgp 65555
 address-family ipv4 unicast
   network 172.16.0.0/16
   redistribute nhrp
 exit-address-family
@end group
@end example


@node Configuring NHRP
@section Configuring NHRP

FIXME

@node Hub Functionality
@section Hub Functionality

In addition to routing nhrp redistributed host prefixes, the hub nodes
are also responsible to send NHRP Traffic Indication messages that
trigger creation of the shortcut tunnels.

nhrpd sends Traffic Indication messages based on network traffic captured
using NFLOG. Typically you want to send Traffic Indications for network
traffic that is routed from gre1 back to gre1 in rate limited manner.
This can be achieved with the following iptables rule.

@example
@group
iptables -A FORWARD -i gre1 -o gre1 \
	-m hashlimit --hashlimit-upto 4/minute --hashlimit-burst 1 \
	--hashlimit-mode srcip,dstip --hashlimit-srcmask 24 --hashlimit-dstmask 24 \
	--hashlimit-name loglimit-0 -j NFLOG --nflog-group 1 --nflog-range 128
@end group
@end example

You can fine tune the src/dstmask according to the prefix lengths you
announce internal, add additional IP range matches, or rate limitation
if needed. However, the above should be good in most cases.

This kernel NFLOG target's nflog-group is configured in global nhrp config
with:
@example
@group
nhrp nflog-group 1
@end group
@end example

To start sending these traffic notices out from hubs, use the nhrp
per-interface directive:
@example
@group
interface gre1
 ip nhrp redirect
@end group
@end example

@node Integration with IKE
@section Integration with IKE

nhrpd needs tight integration with IKE daemon for various reasons.
Currently only strongSwan is supported as IKE daemon.

nhrpd connects to strongSwan using VICI protocol based on UNIX socket
(hardcoded now as /var/run/charon.vici).

strongSwan currently needs few patches applied. Please check out the
@uref{http://git.alpinelinux.org/cgit/user/tteras/strongswan/log/?h=tteras-release,release}
and
@uref{http://git.alpinelinux.org/cgit/user/tteras/strongswan/log/?h=tteras,working tree}
git repositories for the patches.

@node NHRP Events
@section NHRP Events

FIXME

@node Configuration Example
@section Configuration Example

FIXME