[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
(usagi-users 02980) Re: radvd client problems ....
Hi Romain,
Once again thanks for your valuable advice. I have edited the file following
your recomendations. I did so by doing the changes separately and then
recompiling creating two different kernels. Unfortunately when I carried out
the tests I was unable to run IPv6 Fowarding=1 and process RA updates (
accept_ra=1) . I can see the radvd updates, yet the interface does not
process them. Anything else you think I should be aware of ?
Regarding the address allocation, this is my radvd.conf file:
interface eth0
{
AdvSendAdvert on;
prefix fec0:106:2300::0/64
{
AdvOnLink on;
AdvAutonomous on;
AdvRouterAddr on;
};
};
tcpdump:
18:05:27.206461 :: > ff02::1:ff68:b8b0: HBH icmp6: multicast listener report
max resp delay: 0 addr: ff02::1:ff68:b8b0 [hlim 1]
18:05:27.484730 :: > ff02::1:ff68:b8b0: icmp6: neighbor sol: who has
fe80::210:7aff:fe68:b8b0
18:05:28.484758 fe80::210:7aff:fe68:b8b0 > ipv6-allrouters: icmp6: router
solicitation
18:05:30.137859 fe80::20b:dbff:fec3:2de1 > ipv6-allnodes: icmp6: router
advertisement
18:05:30.138007 fe80::210:7aff:fe68:b8b0 > ff02::1:ff07:cae7: HBH icmp6:
multicast listener report max resp delay: 0 addr: ff02::1:ff07:cae7 [hlim 1]
18:05:30.614729 :: > ff02::1:ff68:b8b0: icmp6: neighbor sol: who has
fec0:106:2300:0:210:7aff:fe68:b8b0
18:05:30.954733 :: > ff02::1:ff07:cae7: icmp6: neighbor sol: who has
fec0:106:2300:0:d4a2:8e82:af07:cae7
and the ifconfig data:
eth0 Link encap:Ethernet HWaddr 00:10:7A:68:B8:B0
inet6 addr: fec0:106:2300:0:210:7aff:fe68:b8b0/64 Scope:Site
inet6 addr: fe80::210:7aff:fe68:b8b0/64 Scope:Link
inet6 addr: fec0:106:2300:0:d4a2:8e82:af07:cae7/64 Scope:Site
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1 errors:0 dropped:0 overruns:0 frame:0
TX packets:8 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:100
RX bytes:110 (110.0 b) TX bytes:648 (648.0 b)
Interrupt:11 Base address:0x4000
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:144 errors:0 dropped:0 overruns:0 frame:0
TX packets:144 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:18144 (17.7 Kb) TX bytes:18144 (17.7 Kb)
As you can see I get two "Site" addresses. When I use a DHCPv6 server
connected to the Internet and acting as a Proxy, I get 2-Site addresses +
2-Global addresses. Any hints in where the problem might be ?
Many thanks in advance,
Regards,
Daniel
----- Original Message -----
From: "Romain KUNTZ" <kuntz@xxxxxxxxxxxxxx>
To: <usagi-users@xxxxxxxxxxxxxx>
Sent: Monday, June 28, 2004 1:34 PM
Subject: (usagi-users 02979) Re: radvd client problems ....
> Hi,
>
> > Many thanks indeed for your valuable advice! ... I will perform the
> > changes and recompile the kernel.
>
> Actually I have replaced this line (l. 1076) :
> if (in6_dev->cnf.forwarding || !in6_dev->cnf.accept_ra) {
>
> With this line :
> if ( (in6_dev->cnf.forwarding && in6_dev->cnf.accept_ra < 2)
> || !in6_dev->cnf.accept_ra) {
>
> So if forwarding is enabled, RA are processed only if accept_ra is >1.
> I think this solution is cleaner, because even if forwarding is
> activated, you can choose to accept RA or not for each interface by
> changing the value of the accept_ra flag of the interface (with the
> sysctl command)
>
> > Just one very last thing, when addresses are allocated via radvd and
> > using USAGI's kernel I always get 2-Site-IPv6 addresses on the NIC (
> > this does not happen with SuSE's kernel - only one is assigned ),
> > I have no ideas where to look at, any advice ?
>
> Can I see a dump if the ifconfig command on your NIC please ?
> How many prefixes are advertised on your network ?
>
> Regards,
>
> Romain
>
> > ----- Original Message -----
> > From: "Romain KUNTZ" <kuntz@xxxxxxxxxxxxxx>
> > To: <usagi-users@xxxxxxxxxxxxxx>
> > Sent: Friday, June 25, 2004 5:47 PM
> > Subject: (usagi-users 02976) Re: radvd client problems ....
> >
> >
> > > Hi,
> > >
> > > > I have done some tests with radvd, downloading the original files,
> > > > compiling and running. It seems that my problem is not related to
> > > > radvd. Is it possible on Usagi's Stable Kit 5.0 to support IPv6
> > > > forwarding (sysctl-w net/ipv6/conf/all/forwarding=1) and stateless
> > > > configuration (sysctl-w net/ipv6/conf/all/accept_ra=1 ) at the
> > > > same
> > > > time ? ...
> > >
> > > If forwarding is set to 1, received RA are not processed anymore,
> > > even
> > > if accept_ra is set to one. You can change this behaviour by
> > > modifying
> > > NDP source code :
> > >
> > > /usagi/usagi/kernel/linux26/net/ipv6/ndisc.c (l. 1076)
> > > - if (in6_dev->cnf.forwarding || !in6_dev->cnf.accept_ra) {
> > > + if (!in6_dev->cnf.accept_ra) {
> > >
> > > But it is not recommended that routers accept RAs.
> > >
> > > > everytime I turn on the IPv6 forwarding my NIC no longer sends
> > > > Router
> > > > Solicitations.
> > >
> > > You can still send RS using the rtsol command, but the received RA
> > > will
> > > not be processed, as I told above.
> > >
> > > > However this does not happen with SuSE's native kernel
> > > > 2.4.21-215-default ( I can run sysctl -w
> > > > net/ipv6/conf/all/forwarding=1 and yet send Router Solicitations
> > > > and
> > > > update routing tables). Is there a kernel option I should be aware
> > > > of?
> > >
> > > I guess NDP implementation is not the same.
> > >
> > > Regards,
> > >
> > > --
> > > Romain KUNTZ
> > > kuntz@xxxxxxxxxxxxxx
> > >
> > >
> > >
> >
> >
> >
>
>
> --
> Romain KUNTZ
> kuntz@xxxxxxxxxxxxxx
>
>
>