[nsp] High CPU question

[Bryan]

I have a customer that is running the following config:
7206-nonvxr
npe-225 w/ 256mb ram
I/O-FE
5x PA-Fe-TX interfaces.
IOS 12.2.21b

They are running Ebgp multihop loadbalancing over 2 FE links and pushing
around 40mb/s and the cpu is running about 66% average spiking up to 100%
quite often.  here is an output from the current util:

CPU utilization for five seconds: 25%/24%; one minute: 30%; five minutes:
26%

when it is experiencing problems it show's 66%/66% which indicates that
the interrupts are causing the high utilization but there is not
indication that there is anything wrong, according to the troubleshooting
steps on CCO.

Here is the BGP config:

router bgp 1234
 bgp log-neighbor-changes
 network 10.0.100.0 mask 255.255.254.0
 network 10.0.155.0 mask 255.255.255.0
 network 10.0.188.0 mask 255.255.255.0
 redistribute connected
 redistribute static
 neighbor 100.0.255.158 remote-as 11111
 neighbor 100.0.255.158 ebgp-multihop 2
 neighbor 100.0.255.158 password 7
 neighbor 100.0.255.158 update-source Loopback0
 neighbor 100.0.255.158 version 4
 neighbor 100.0.255.158 send-community
 neighbor 100.0.255.158 remove-private-AS
 neighbor 100.0.255.158 distribute-list BGP-TO-PROVIDER out
 neighbor 100.0.255.158 route-map NSP-OUT out
 no auto-summary

static routes for other end loop back poitn to their direct connect
interfaces and static routes for teh 3 networks they are trying to
advertise pointed to other loopback for loadbalancing.

I've turned off all of the processes I can think of and have CEF turned on
try and speed things up and take some CPE utilization away.


Any thoughts as to what could be causing this?  Is it a bad processor, MD5
password causing it?

I'm pretty much at a loss as to what i could be so any help is greatly
appreciated.




TIA,



Bryan