RE: Why pim dm-fallback ???

Aly,

This was discussed last week, for practical lab exercises, it is not really needed. I wouldn't do the explanation justice, but it is valid to enable dm-fallback when using sparse-mode only.

This is from Mohammed on the 29th:

Totally agree with you, Cisco documentation regarding this point is a little confusing.

I was in this dilemma before, your understanding is correct (IMHO), as long as you are not using sparse-dense interfaces thus the "no ip pim dm-fallback" command is of no need (but please read the email to its end), try labing it out, moreover to practically prove this, when I tried disabling and enabling this command on a router with all interfaces in sparse mode I get the following (the router is a Cisco 3640 running the
12.3(14)T7 code - using dynamips):

Rack1R3(config)#ip pim dm-fallback
Can not have fallback mode as dense, all interfaces are in pim sparse mode

Moreover in the "PIM Dense Mode Fallback Prevention in a Network Following RP Information Loss" document:
http://www.cisco.com/en/US/docs/ios/12_3t/12_3t4/feature/guide/gtautorp.html#wp1045491

I quote "By default, if all interfaces in a multicast VPN routing or forwarding instance are configured with the ip pim sparse-mode command, there is no dense mode fallback because dense mode groups cannot be formed over interfaces configured for sparse mode."

Although within the Cisco documentation command reference for this command:
http://www.cisco.com/en/US/docs/ios/ipmulti/command/reference/imc_04.html#wp1013545

I quote "PIM dense mode fallback is enabled. That is, a multicast group in the absence of rendezvous point (RP) information will fall to dense mode, regardless of the interface mode configuration."

My opinion is that the answer lies in this paragraph from the first
document:
http://www.cisco.com/en/US/docs/ios/12_3t/12_3t4/feature/guide/gtautorp.html#wp1050843

"By default, PIM dense mode fallback is enabled. That is, a multicast group in the absence of rendezvous point (RP) information will fall to dense mode, regardless of the interface mode configuration.

However, if all of the interfaces in a VRF are configured with PIM sparse mode, no dense mode fallback is achieved by default. If all of the interfaces are already configured as sparse, even though the group mode falls to dense mode, the traffic does not get flooded (due to the sparse characteristic of the interface). But the established flows might be torn down and the state of the network could become indeterministic. The main advantage of no dense mode fallback in this case would be deterministic behavior. "

To recap, this means that with all interfaces in sparse mode, although the group mode will fall to dense mode, the traffic can't be forwarded in dense mode over the sparse mode interfaces, and thus there will be no traffic flow, and this might introduce indeterministic behavior and here comes the benefit of this command in this case.

Practical labing does the proof of concept for you, so stick with it, and try reading more than one document and verifying their contents.

Have a nice day.

BR,
Mohammed Mahmoud.

-ryan

-----Original Message-----
From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On Behalf Of aly.groupstudy_at_gmail.com
Sent: Monday, May 04, 2009 10:42 AM
To: ccielab_at_groupstudy.com
Subject: Why pim dm-fallback ???

What is the exact difference between using the following
!1
Router(config)#no ip pim dm-fallback
Router(config-if)#ip pim sparse-dense-mode
--------------------------------------------
--------------------------------------------
!2
Router(config-if)#ip pim sparse-mode

I really could not find any difference and cannot understand the logic of cisco
behind having this command "no ip pim dm-fallback"

I would really appreciate your feedback!

Blogs and organic groups at http://www.ccie.net
Received on Mon May 04 2009 - 10:49:35 ART