Our ASA [ver. 9.10(1)] does multicast routing and the source is directly connected on inside. Every once in awhile a multicast stream (group) does not get forwarded to the RP (outside) and is dropped (show mfib ), and stays dropped. Performing a clear pim topology for the group address helps it along. (Haven't tested with clear arp for the address). The drop reason (from capture) is that the Punt rate limit exceeded. I am assuming that the multicast streams hitting the inside interface is causing the excessive arp entries to exceed the 500 arp rate limit (per interface). I am guessing that some streams are forwarded normally ( and do not "get stuck") because the SPT tree is built in a timely manner for connection to flow.
Right now the rate limit is defaulted to 32768. Would adjusting this to below 500 for the inside interface fix the problem ? or is there something else for a better fix. I'm surprise that the device can't handle the stream amounts until a proper join to the RP has been achieved.
I also read that the punt rate limit was a bug as well, but our version is no in that category.
