All things related to multicast

["Brown, Robin" <>]

Yes, we can vouch for this.  Higher than normal CPU on routers in the
path where there was obviously a SAP listener downstream.  It also
invloved a lot of PIM traffic to our RP, this was also an impact on CPU
for us. I think though that not having sap listen enabled on our routers
kept us from total meltdown.

We were fortunate this time that it was not causing service disruption.
We have sap listen disabled.

Group: 224.2.127.254, (SAP.MCAST.NET)
   Source: 81.194.43.1 (?)
     Rate: 16711 pps/43716 kbps(1sec), 43737 kbps(last 50 secs), 32
kbps(life avg)

We're still considering blackhole-ing this mroute.

-Robin


-----Original Message-----
From: Caren Litvanyi 
[mailto:]
 
Sent: Wednesday, June 11, 2008 3:33 PM
To: 

Subject: RE: Another SAP Storm?

Hi,
I have been busy passing around info how to turn off SAP listen on
Junipers and Ciscos, and have been pretty sure that will mitigate all
high CPU issues.
Our IU engineers replied that is not the case, that additionally if
someone routed off of a given router joins the SAP group, it also sends
their cpu sky-high.
Is anyone else experiencing the same high CPU effects even though SAP
listen is completely off on the router?
I don't see how this would be different than any other group.
Thanks,
Caren
------------
Caren Litvanyi, network engineer
Internet2 NOC
Global Research NOC at Indiana University 

o:812-961-3790  c:250-896-4369