dude can cause router or switch cpu Instantaneous peak high to 95% about , I change polling interval to 2min ,label refresh interva to 24 hour,but cpu peak still very high ,when I stop dude ,all cpu peak became normal,cisco doc said ,snmp polling on arp table or route table may cause this . the service I polling only ping，label is like devicename，ip addr，link label is interface rx and tx speed. anywhere can I setup dude don't polling the route table and arp table.
Very urgent,anyone can solve this problem!

I'm not having this exact issue, but would also like to know if I can only poll SNMP x, y or z (Ie, just Interfaces). Some of the route/arp/bridgefdb tables on some of the Wireless devices I have are huge, and I can see it causing a problem in the future.

Maybe you can disable some snmp parameters on cisco routers(don't know can it be done on cisco or not)
If you only need ping, than create fake snmp profile, or disable snmp read at all for dude ip address

I have run into this problem not only with the dude, but with many network management systems that probe all the interfaces on a router/switch. The problem comes from if you have a high number or arp addresses or interfaces/subinterfaces that causes the processor on the device to go through the roof. The best solution for this is to limit the number of interfaces displayed or oids that can be walked. Yes this is something that can be done on Cisco devices fairly easily. The "snmp-server view cutdown" command will limit different snmp values. Here is a link that explains this problem http://www.cisco.com/en/US/tech/tk648/t ... 48e6.shtml