Skip to Content.
Sympa Menu

wg-multicast - Yet another broken multicast at Columbia

Subject: All things related to multicast

List archive

Yet another broken multicast at Columbia


Chronological Thread 
  • From: Alan Crosswell <>
  • To:
  • Subject: Yet another broken multicast at Columbia
  • Date: Wed, 8 Aug 2001 22:35:47 EDT

[I finally fixed my nlanr/AG beacon and went to test it and discovered
multicast had broken on my subnet:-].

So this time, we have a router that has one of it's two redundant
links to the rest of the campus routers down (dual star topology;
subnets 1 and 2). All campus routers with the exception of the I2
router (and RP) are on subnets 1 and 2. The I2 router is only on
subnet 1. The router that has a link down has subnet 1 down. Unicast
routing works just fine via subnet 2 where one of the other half-dozen
routers moves the subject router's traffic from subnet 2 back to
subnet 1 for access to the I2 router/RP.

We are using auto-rp. I do a "show ip pim rp" successively and see
the RP present for a very small number of groups, and then a second
later no RPs, and then a while later RPs but again only for a small
number of groups. I have some debug output that I'll send to TAC
RSN....

The workaround was to manually configure the I2 router as the RP (ip
pim rp-address 128.59.0.6) and multicast connectivity is back.

So I guess it's something funky going on with the way auto-rp works which
I thought was PIM-DM which is why the router interfaces have to be configured
in sparse-dense mode.

IOS version is:

IOS (tm) c6sup2_rp Software (c6sup2_rp-JSV-M), Experimental Version
12.1(20010728:011232) [shj-V121_8A_E 109]

It's stable, really! It was built a whole week ago:-)

/a

PS: This time when Phil at Cornell replies saying how he solved it I
will pay attention and not just save the mail for 6 months!




Archive powered by MHonArc 2.6.16.

Top of Page