Browse code

4a, 9, 10, 11, 12 added - and 11. done right away :-) (cherry picked from commit ea382a1d550ac100d27c8118777e3160c85d06d2)

Gert Doering authored on 2010/01/14 23:52:06
Showing 1 changed files
... ...
@@ -22,6 +22,17 @@ known issues for IPv6 payload support in OpenVPN
22 22
     For Solaris, only the "ipv6 tun0" is affected, for the *BSDs all tun0
23 23
     stay around.
24 24
 
25
+4a.) deconfigure IPv6 on tun interface on session termination, otherwise
26
+    one could end up with something like this (on NetBSD):
27
+
28
+tun0: flags=8051<UP,POINTOPOINT,RUNNING,MULTICAST> mtu 1500
29
+        inet 10.9.0.18 -> 10.9.0.17 netmask 0xffffffff
30
+        inet6 fe80::a00:20ff:fece:d299%tun0 ->  prefixlen 64 scopeid 0x3
31
+        inet6 2001:608:4:eff::2000:3 ->  prefixlen 64
32
+        inet6 2001:608:4:eff::1:3 ->  prefixlen 64
33
+
34
+    (pool was changed, previous address still active on tun0, breakage)
35
+
25 36
 5.) add new option "ifconfig-ipv6-push"
26 37
     (per-client static IPv6 assignment, -> radiusplugin, etc)
27 38
 
... ...
@@ -35,3 +46,26 @@ known issues for IPv6 payload support in OpenVPN
35 35
 
36 36
 8.) full IPv6 support for TAP interfaces 
37 37
     (main issue should be routes+gateway - and testing :-) )
38
+
39
+9.) verify that iroute-ipv6 and route-ipv6 interact in the same way as
40
+    documented for iroute/route:
41
+
42
+    A's subnet, OpenVPN must push this route to all clients
43
+    EXCEPT for A, since the subnet is already owned by A.
44
+    OpenVPN accomplishes this by not
45
+    not pushing a route to a client
46
+    if it matches one of the client's iroutes.
47
+
48
+10.) extend "ifconfig-ipv6" to handle specification of /netbits, pushing
49
+    of /netbits, and correctly ifconfig'ing this
50
+    (default, if not specified: /64)
51
+
52
+11.) do not add ipv6-routes if tun-ipv6 is not set - complain instead
53
+
54
+     * done * 12.1.10
55
+
56
+12.) handle incoming [::] and [fe80:...] packets in tun-p2mp MULTI mode
57
+     (most likely those are DAD packets)
58
+     silently ignore DAD?  
59
+        Or accept-and-forward iff (multicast && client2client)?
60
+     handle NS/NA