[c-nsp] High CPU caused by interupt on 7600 router
Rin
rintrum at gmail.com
Fri Oct 1 11:24:06 EDT 2010
Hi Phil,
Thanks for your email. Below is my answer:
> Do you have any CoPP or MLS limiters enabled? I am thinking particularly
the TTL and MTU ones.
The CPU only spikes in less than a minute show I cannot catch it when it
high. I had CoPP applied as below
==============================================================
ip access-list standard CoreIP
permit 172.16.x.x 0.0.0.255
permit 172.16.x.x 0.0.0.255
!
class-map match-all CoreIP
match access-group name CoreIP
!
policy-map CoPP
class CoreIP
police 20000000 conform-action transmit exceed-action drop
class class-default
police 6000000 conform-action transmit exceed-action drop
!
control-plane
service-policy input CoPP
==============================================================
The output shows more than 12Mbps of traffic (matched by class-default) is
sending to CPU. I believe this is value is high but I could not determine
what type of traffic is sending to CPU
============================================================================
===
PE-Router#sho policy-map control-plane
Control Plane
Service-policy input: CoPP
Hardware Counters:
class-map: CoreIP (match-all)
Match: access-group name CoreIP
police :
20000000 bps 625000 limit 625000 extended limit
Earl in slot 1 :
1684255 bytes
5 minute offered rate 2808 bps
aggregate-forwarded 1684255 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 3760 bps exceed 0 bps
Earl in slot 4 :
0 bytes
5 minute offered rate 0 bps
aggregate-forwarded 0 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Earl in slot 6 :
0 bytes
5 minute offered rate 0 bps
aggregate-forwarded 0 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Earl in slot 7 :
0 bytes
5 minute offered rate 0 bps
aggregate-forwarded 0 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Earl in slot 9 :
0 bytes
5 minute offered rate 0 bps
aggregate-forwarded 0 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Software Counters:
Class-map: CoreIP (match-all)
27872 packets, 2226445 bytes
5 minute offered rate 4000 bps, drop rate 0000 bps
Match: access-group name CoreIP
police:
cir 20000000 bps, bc 625000 bytes
conformed 27915 packets, 2229727 bytes; actions:
transmit
exceeded 0 packets, 0 bytes; actions:
drop
conformed 4000 bps, exceed 0000 bps
Hardware Counters:
class-map: class-default (match-any)
Match: any
police :
6000000 bps 187000 limit 187000 extended limit
Earl in slot 1 :
7697842499 bytes
5 minute offered rate 12430840 bps
aggregate-forwarded 3726798935 bytes action: transmit
exceeded 3971043564 bytes action: drop
aggregate-forward 6016104 bps exceed 6414120 bps
Earl in slot 4 :
0 bytes
5 minute offered rate 0 bps
aggregate-forwarded 0 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Earl in slot 6 :
0 bytes
5 minute offered rate 0 bps
aggregate-forwarded 0 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Earl in slot 7 :
908447 bytes
5 minute offered rate 880 bps
aggregate-forwarded 908447 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 624 bps exceed 0 bps
Earl in slot 9 :
304 bytes
5 minute offered rate 0 bps
aggregate-forwarded 304 bytes action: transmit
exceeded 0 bytes action: drop
aggregate-forward 0 bps exceed 0 bps
Software Counters:
Class-map: class-default (match-any)
7093 packets, 1164410 bytes
5 minute offered rate 2000 bps, drop rate 0000 bps
Match: any
7093 packets, 1164410 bytes
5 minute rate 2000 bps
police:
cir 6000000 bps, bc 187500 bytes
conformed 7101 packets, 1165380 bytes; actions:
transmit
exceeded 0 packets, 0 bytes; actions:
drop
conformed 3000 bps, exceed 0000 bps
============================================================================
===
> You could ERSPAN the RP/SP CPU to a remote machine, then correlate CPU
spikes with the captured traffic.
How can I do this? Do you have any guideline?
> How many BGP routes and peers do you have? What is the churn rate on the
BGP table? Is there any IGP stability? Is there any possibility of a loop?
We only run MP-BGP for MPLS VPN on this router. The BGP table contains less
than 10 routes. I do not tune any BGP timer and keep it as default. We run
OSPF inside the network and seems there is no instability.
Thanks,
Rin
More information about the cisco-nsp
mailing list