Re: A multicast confusion!!

From: oo IPX <oispxl_at_gmail.com>
Date: Fri, 21 Sep 2012 18:19:37 +0530

Hi Carlos,

The ascii, It got changed, I guess group-study modifies things before
posting.
I have uploaded the same on the following link, and "have pasted the
question below" just to save scrolling.

--- https://docs.google.com/open?id=0B2uwG9h65O8BQVBSQVpqb0Y2U1k ---

I tried to verify myself but the extensive ouput of debugs prompted to ask
:-)

The scenario is as shown above, R4 is *statically *configured RP, on all
the routers ('not on R6' and can't remember about R3 )
Now there is sth i forgot ( R3 is either not running ip multicast or the
pim sparse is not configured on interfaces - i can't remember)

the underlying unicast routing is such that R1 chooses link through R2-R6
to reach R5, and R2-R4 to reach RP.

This is what I observed, and just want to know if its correct.

1. R5 sources the ping to 224.1.1.1
2. R5 sends unicast pim register message to R4,
3. consider R1 is already configured to receive that group.
4. R5 starts forwarding de-capsulated packets to R2-->R1 --- R6 is a pim
neighbor but not running multicast.
5. R1 sends rpbit set prune, for the (*,G) towards R1.

6. couldn't decipher whether R1 forwards that (8,G) prune to RP or not,
but as yet the RP is still forwaring decapsulaed packets.

7. configure multicast support on R6.

8. But, R6 have no information about RP as yet, so i added the command ip
pim rp-address R5.

9. Now starts the doubt??

does the R6 sends (*,G) join to RP, and once it get the source address for
the G, it prunes from (*,G) and builds SPT toward the source.??

also the command 'show ip mroute count" shows that packet are indeed being
forwarded out to R2 for SPT afterwards.

However at this time the RP starts sending Register-Stop messages to source
which were not being sent earlier.

I might be confusing myself and the reader.
 Your views are greatly appreciated.
Thanks.

On Fri, Sep 21, 2012 at 4:19 PM, Carlos G Mendioroz <tron_at_huapi.ba.ar>wrote:

> oo,
> your ascii art got really confusing... you might want to try it again.
>
>
>
> oo IPX @ 20/09/2012 17:39 -0300 dixit:
>
>> R1 (receiver) -------------R2-------------***
>> R4*-(*RP*)-------R3==========**R5 *
>> * \
>> *(source)*
>>
>> \ /
>>
>> \___________R6______________/
>>
>
> --
> Carlos G Mendioroz <tron_at_huapi.ba.ar> LW7 EQI Argentina

Blogs and organic groups at http://www.ccie.net
Received on Fri Sep 21 2012 - 18:19:37 ART

This archive was generated by hypermail 2.2.0 : Mon Oct 01 2012 - 06:40:29 ART