These commands are helpful to identify causes of high CPU on 6500's with sup720.
Cisco.com doc below as well.
Check CPU levels
## sh process cpu | ex 0.00
## sh process cpu history
Identify spanning tree specific information. Will show us rx and tx
counters and forwarding transitions.
## show spanning-tree detail | inc ieee|occur|from|is exec
This command will give a detailed snapshot for STP showing if any vlans have had recent TCN's. If a vlan has a lot
of TCN's for spanning tree, that can cause state changes and can cause stability issues including flooding to
VLAN0001 is executing the rstp compatible Spanning Tree protocol
Number of topology changes 19 last change occurred 00:05:26 ago
Port 4096 (port-channel1, vPC Peer-link) of VLAN0001 is designated forwarding
In this case there was a topology change in the last 5 minutes. If it was occuring every minute or less
there is something to focus on.
##sh spanning-tree summary
##sh spanning-tree vlan [vlan id] detail
Check interface statistics for problems
##sh interface | inc rate|line
This command is useful for identifying interfaces with high rates of traffic. If you find an interface that is extremely high with with traffic that normally may not exist, you can shut down that interface or find out what is on the other end. The end device might be sending traffic that the switch cpu is needing to process resulting in high CPU.
GigabitEthernet6/1 is up line protocol is up
Queueing strategy: fifo
5 minute input rate 900010000 bits/sec, 514 packets/sec
5 minute output rate 1200 bits/sec, 3 packets/sec
##sh interface counters errors
##sh int | in Input|line
##sh interface switching
Check logs on both RP and SP.
##remote command show log
Check the health of the inband port going to the cpu.
##remote command switch show ibc
Have a layer 2 topology ready to view to confirm links. Double
confirm using CDP if enabled.
##sh cdp neighbors detail
## Visio layer 2 topology
7: Span traffic heading to the cpu.
## Use netdr debug tool.
debug netdr capture rx
show netdr capture