Hi,
In our network I am seeing high CPU up to more then 90% when our multicast streams are enabled. CPU drops instantly to normal (20-30%) when I turn off the streams for a minute.
a HSRP redundant L3 switch named switch2 (Catalyst 3850) was recently added to our network which is configured as standby router. See also network diagram below. I notice that specifically on this device the CPU usage is very high while no other traffic passes the device because it is the standby router and all other links to the switch are blocked by spanning-tree.
PIM anycast RP is configured on both switches and sparse-mode is configured on the SVI interfaces. IPTV server is on vlan 250 and clients are on multiple vlans but have choosen vlan 170 for the client now to keep things simple. I can recieve the iptv streams on the client just fine.
I added some config files aswel in attachment which show the config of the interfaces, pim config, cpu processes, mroutes, ...
I believe switch2 is also functioning as the RP because switch1 is the DR. I checked the traffic on the inter switch connection and I can see all the streams on this link. I would expect that pim would prune them because there are no clients connected?
I have spent quite some time on this issue also looking on this forum and the cisco documentation without any luck. I am wondering if this behaviour is expected or something is still wrong with my config?
Any help is highly appreciated
