Re: multicast packets

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



>>>>>>>>>>>>Looking a little bit in the source-code it seems that it installs a route

True . it installs multicast route  { /prox/net/ip_mr_cache }.
But route comes into play only at POSTROUTING hook . { after reaching
box } .Since i dont have any rule in INPUT to accept those ,it should
be rejected .

You confirmed that IGMP wont go to ESTABLISHED state . I think , i
have to debug more .



On Fri, Feb 26, 2010 at 12:50 PM, Christoph Paasch
<christoph.paasch@xxxxxxxxx> wrote:
> I don't know the details about igmpproxy.
> Looking a little bit in the source-code it seems that it installs a route
> inside in the kernel for a certain multicast-stream.
> Thus, multicast-udp will pass your gateway as they don't hit the INPUT chain.
>
> However for the igmp-control traffic I have no idea, why igmpproxy receives
> these.
>
> Christoph
>
> On Fri 26 February 2010, ratheesh k wrote:
>> I am running  igmp stream  client to stream igmp packets on machine A
>> from internet .I can play files .
>>
>> note : Is igmpproxy running machine B has anything to do with this  ?
>>
>>
>>
>> On Fri, Feb 26, 2010 at 12:11 PM, Christoph Paasch
>>
>> <christoph.paasch@xxxxxxxxx> wrote:
>> > On Fri 26 February 2010, ratheesh k wrote:
>> >> INPUT  policy is DROP
>> >> FORWARD policy is DROP
>> >> OUTPUT policy is accept
>> >>
>> >>
>> >> INPUT chain
>> >> iptables -A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT
>> >> iptables -A INPUT  -i eth0  -j ACCEPT
>> >>
>> >>
>> >> FORWARD
>> >> iptables -A FORWARD  -m state --state ESTABLISHED,RELATED -j ACCEPT
>> >> iptables -A INPUT  -i eth0  -o eth1 -j ACCEPT
>> >
>> > I imagine that this is a typo and that you meant the FORWARD chain.
>> >
>> >> machine                    Gateway machine B
>> >>    A -------------------->eth0              eth1 -------------> internet
>> >>
>> >> I have a machine A . which is connected to a linux gateway machine .
>> >> Ruleset and policy mentioned are for machine B . There is no iptables
>> >> rules in machine A .
>> >>
>> >> >>>>>>>>>What do you want to achieve ?
>> >>
>> >> As per the current rule , no igmp packet should come to GATEWAY
>> >> machine , since there is no firewall hole in input chain .
>> >>
>> >> >>>>>>>>>>what are you observing?
>> >>
>> >> But i can see , igmp packets flowing into machine B from internet .
>> >
>> > Where can you see these igmp packets? With wireshark/tcpdump? If it is
>> > one of those, than it is normal, because these capture the packets
>> > before the iptables filter.
>> >
>> > Seen your ruleset, the packets should not enter, if they are coming from
>> > the internet.
>> >
>> > Regards,
>> > Christoph
>> >
>> >> On Fri, Feb 26, 2010 at 4:46 AM, Christoph Paasch
>> >>
>> >> <christoph.paasch@xxxxxxxxx> wrote:
>> >> > Please, provide more information about your setup.
>> >> >
>> >> > What are the policies of your chains? What is your ruleset?
>> >> > What is your topology?
>> >> > What do you want to achieve, and what are you observing?
>> >> >
>> >> > Christoph
>> >> >
>> >> > On Thu 25 February 2010 wrote ratheesh k:
>> >> >> iptables -A INPUT -m state --state ESTABLISHED,RELATES -j ACCEPT .
>> >> >>
>> >> >> This is the only rule . No firewall hole for igmp packets .
>> >> >>
>> >> >> On Thu, Feb 25, 2010 at 12:08 PM, ratheesh k <ratheesh.ksz@xxxxxxxxx>
>> >
>> > wrote:
>> >> >> >>>>>>>>>>udp doesn't go into the established state.
>> >> >> >
>> >> >> > I am running "igmpproxy" on my gateway box . I didnot add any rule
>> >> >> > in INPUT chain to accept igmp packets . But  i hve a rule to
>> >> >> > accept all ESTABLISHED state packets . It am able to stream igmp
>> >> >> > from my desktop .
>> >> >> >
>> >> >> > I really believe that " We dont need any rule in FORWARD chain " .
>> >> >> > Because packets are flowing from node to node and routed . So only
>> >> >> > INPUT and OUTPUT chains are involved .
>> >> >> >
>> >> >> > Thanks,
>> >> >> > Ratheesh
>> >> >> >
>> >> >> >
>> >> >> >
>> >> >> > On Thu, Feb 25, 2010 at 12:03 AM, Christoph Paasch
>> >> >> >
>> >> >> > <christoph.paasch@xxxxxxxxx> wrote:
>> >> >> >> As long as there isn't any return-traffic (as it is the case for
>> >> >> >> multicast- udp), udp doesn't go into the established state.
>> >> >> >>
>> >> >> >> Regards,
>> >> >> >> Christoph
>> >> >> >>
>> >> >> >> On Wed 24 February 2010 wrote ratheesh k:
>> >> >> >>> multicast packets are udp packets . But its flowing only from
>> >> >> >>> upstream to downstream . So packet state will be always "NEW" .
>> >> >> >>> ??
>> >> >> >>>
>> >> >> >>> my question is : whether we can see multicast data packets in
>> >> >> >>> ESTABLISHED state ??
>> >> >> >>>
>> >> >> >>> Thanks,
>> >> >> >>> Ratheesh
>> >> >> >>> --
>> >> >> >>> To unsubscribe from this list: send the line "unsubscribe
>> >> >> >>> netfilter" in the body of a message to majordomo@xxxxxxxxxxxxxxx
>> >> >> >>> More majordomo info at
>> >> >> >>>  http://vger.kernel.org/majordomo-info.html
>> >> >> >>
>> >> >> >> --
>> >> >> >> Christoph Paasch
>> >> >> >>
>> >> >> >> Alcatel-Lucent
>> >> >> >> IP Development
>> >> >> >>
>> >> >> >> www.rollerbulls.be
>> >> >> >> --
>> >> >> >> --
>> >> >> >> To unsubscribe from this list: send the line "unsubscribe
>> >> >> >> netfilter" in the body of a message to majordomo@xxxxxxxxxxxxxxx
>> >> >> >> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> >> >>
>> >> >> --
>> >> >> To unsubscribe from this list: send the line "unsubscribe netfilter"
>> >> >> in the body of a message to majordomo@xxxxxxxxxxxxxxx
>> >> >> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> >> >
>> >> > --
>> >> > Christoph Paasch
>> >> >
>> >> > Alcatel-Lucent
>> >> > IP Development
>> >> >
>> >> > www.rollerbulls.be
>> >> > --
>> >>
>> >> --
>> >> To unsubscribe from this list: send the line "unsubscribe netfilter" in
>> >> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> >> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> >
>> > --
>> > Christoph Paasch
>> >
>> > Alcatel-Lucent
>> > IP Development
>> >
>> > www.rollerbulls.be
>> > --
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe netfilter" in
>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> --
> Christoph Paasch
>
> Alcatel-Lucent
> IP Development
>
> www.rollerbulls.be
> --
>
--
To unsubscribe from this list: send the line "unsubscribe netfilter" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Linux Netfilter Development]     [Linux Kernel Networking Development]     [Netem]     [Berkeley Packet Filter]     [Linux Kernel Development]     [Advanced Routing & Traffice Control]     [Bugtraq]

  Powered by Linux