[hipl-users] Re: IPv6 address in LOCATOR parameter and mobility
- From: Miika Komu <miika.komu@xxxxxxx>
- To: hipl-users@xxxxxxxxxxxxx
- Date: Fri, 26 Jun 2009 14:03:23 +0300
Leo Martinez wrote:
Hi Leo,
sorry for long response time, I've got a long queue of tasks...
Hello everyone,
I've being doing some tests of mobility and the UPDATE messages on HIPL
and I have a question regarding the format and of the address field in
the LOCATOR parameter, since it does not represent the addresses
available on the interfaces of the initiators and responders involved in
my tests (I'm using wireshark and the patch procedure provided by samu
in the source code to view the traces).
Are you using the latest version of the wireshark patches? See my
previous email.
Also, I've noticed that the
nodes involved simply won't send the UPDATE messages for ECHO_REQUEST
and ECHO_RESPONSE to the corresponding IPv6 global addresses used during
the BE and the conntest tests (chapter 10 of the manual), the UPDATE
messages are sent to the IPv4 addresses of the nodes.
* Is this behaviour (sending UPDATE messages for ECHO_REQUEST and
ECHO_RESPONSE by the IPv4 addresses ONLY) normal in the hipl
implementation?
No.
* Does the address given by the locator (hipcong locator get) is
always an IPv4 address?
No.
* Chapter 12 (Testing Handover) of the manual suggests to /"Make
sure to add the route for the new address as soon as the new
address is added. If the route is missing, the update packets
might not get sent at all. The daemon waits for a couple of
seconds if a transmit of an update fails for restransmitting it."/
How can this be achieved?
If you add an IPv4 address, the kernel doesn't necessarily any route for
it. You can add the route with "ip route" or "route" commands.
After the UPDATE messages are exchanged, I check the SAs (hipconf get ha
all) and noticed the IP addresses changed from the original global IPv6
to the IPv4 addresses of the nodes and after that, no more messages
reach the conntest-server.
I tried the following:
hipconf add map HIT 3ffe::1
ping6 HIT
ip addr del 3ffe::2/64 dev wlan0 # delete local ipv6 address
ip addr add 3ffe::3/64 dev wlan0
This didn't involve any IPv4 addressing for mobility, albeit I noticed
another bug which I plan to fix.
I'd appreciate your comments based on your experiences guys in order to
understand the locator parameter, why there is no exchange of data
between nodes when the locator parameter is on (hipconf locator on) and
why I can't seem to achieve mobility.
Turn locator off to avoid introducing IPv4 addresses for HIP if you want
to just to play with IPv6. The locator parameter is an unstable feature
and it is therefore disabled by default.
Btw, Baris is improving the mobility code. I'll merge new features from
his branch to the main stream as they get more mature. I assume that
this will occur in August.
- I'm using kernel 2.6.28-11-generic
- the hipd_config file has only the following options uncommented:
* hit-to-ip off
* nsupdate off
* opendht off
* heartbeat 0
* locator on
* nat none
* debug all
- The are the iptables configured in the node:
iptables -L -n
Chain INPUT (policy ACCEPT)
target prot opt source destination
HIPFW-INPUT all -- 0.0.0.0/0 <http://0.0.0.0/0> 0.0.0.0/0
<http://0.0.0.0/0>
Chain FORWARD (policy ACCEPT)
target prot opt source destination
HIPFW-FORWARD all -- 0.0.0.0/0 <http://0.0.0.0/0>
0.0.0.0/0 <http://0.0.0.0/0>
Chain OUTPUT (policy ACCEPT)
target prot opt source destination
HIPFW-OUTPUT all -- 0.0.0.0/0 <http://0.0.0.0/0> 0.0.0.0/0
<http://0.0.0.0/0>
Chain HIPFW-FORWARD (1 references)
target prot opt source destination
Chain HIPFW-INPUT (1 references)
target prot opt source destination
Chain HIPFW-OUTPUT (1 references)
target prot opt source destination
QUEUE all -- 0.0.0.0/0 <http://0.0.0.0/0> 1.0.0.0/8
<http://1.0.0.0/8>
- And the ip6tables configured are:
Chain INPUT (policy ACCEPT)
target prot opt source destination
HIPFW-INPUT all ::/0 ::/0
Chain FORWARD (policy ACCEPT)
target prot opt source destination
HIPFW-FORWARD all ::/0 ::/0
Chain OUTPUT (policy ACCEPT)
target prot opt source destination
HIPFW-OUTPUT all ::/0 ::/0
Chain HIPFW-FORWARD (1 references)
target prot opt source destination
Chain HIPFW-INPUT (1 references)
target prot opt source destination
QUEUE all ::/0 2001:10::/28
Chain HIPFW-OUTPUT (1 references)
target prot opt source destination
- Meanwhile, the only processes I have runing for hip are (ps aux |
grep hip)
nobody 5955 0.0 0.0 13776 1560 pts/1 S+ 16:55 0:00 hipfw
-klpAF
nobody 11656 0.0 0.0 14668 2560 pts/2 S+ 17:45 0:00 hipd -k
- No dns proxies are active (ps aux | grep dns)
root 14204 0.0 0.0 3336 804 pts/5 S+ 18:08 0:00 grep dns
Other related posts:
