[j-nsp] IPv6 multicast -- stuck in spt-pending

Josef Buchsteiner josefb at juniper.net
Mon Feb 28 01:40:34 EST 2005


       once  the  shared  tree  is  equal  to  the spt we will stay in
       spt-pending  state which is perfectly ok and not a bug. PR55948
       is a rpd assertion and somewhat totally different from what you
       describe here.


Monday, February 28, 2005, 12:11:02 AM, you wrote:

DR> Hi,

DR>  Uplink ---Tunnel--- R1 --- R2 --- R3 --- R4 --- R5 --- Host

DR>  R1+R2: Juniper
DR>  R3-R5: Cisco

DR>  Host sends MLDv1 (*,G) join towards DR R5. R5 sends PIM (*,G) join
DR>  to RP. This join is forwarded from R5 to R1 and further. R1+R2 now
DR>  show:

DR>  Group: ff3e:20:2001:660::beac
DR>      Source: *
DR>      RP: 2001:660:3007:300:1::
DR>      Flags: sparse,rptree,wildcard
DR>      Upstream interface: ip-1/3/0.1

DR>  Uplink starts sending multicast traffic on this group down the tunnel
DR>  to R1, which gets forwarded down to R5. R5 then joins each detected
DR>  source with PIM (S,G) joins. Those get forwarded again up to the Uplink.

DR>  At this point, both Junipers R1 and R2 do show:

DR>  Group: ff3e:20:2001:660::beac
DR>      Source: 2001:468:901:1:250:daff:fe20:5b03
DR>      Flags: sparse,spt-pending
DR>      Upstream interface: ip-1/3/0.1

DR>  for each (S,G). Traffic is STILL coming in, and RPT == SPT, so R1 and R2
DR>  should _immediately_ conclude that they are active on the SPT and should
DR>  change state from "spt-pending" to "spt". Unfortunately that doesn't
DR>  happen.

DR>  Smells like a bug to me. Tried with 7.1R1 and 7.0R2.

DR>  This group has some 14 sources, so am I triggering PR 55948?

DR>  Best regards,
DR>  Daniel

DR>  --
DR>  CLUE-RIPE -- Jabber: dr at cluenet.de -- dr at IRCnet -- PGP: 0xA85C8AA0
DR>  _______________________________________________
DR>  juniper-nsp mailing list juniper-nsp at puck.nether.net
DR> http://puck.nether.net/mailman/listinfo/juniper-nsp


More information about the juniper-nsp mailing list