6509 sup720 high cpu

berkcom · ‎02-17-2008

I have a pair of 6509's each having a single sup720-3bxl, 6408 8port gbic, and 2 6748 modules. I am experiencing higher than expected interrupt cpu usage with about 900mbps of inbound ip traffic. I don't have dynamic routing, just a few routed vlans and a couple static routes. I am also doing hsrp for the vlan interfaces between the two. Here's some stats:

show proc cpu | exclude 0.00 :

CPU utilization for five seconds: 40%/39%; one minute: 43%; five minutes: 44%

PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process

122 72629332 359004276 202 0.87% 0.64% 0.67% 0 IP Input

show mls stat :

Statistics for Earl in Module 5

L2 Forwarding Engine

Total packets Switched : 561152787313

L3 Forwarding Engine

Total packets L3 Switched : 559997750297 @ 179965 pps

Total Packets Bridged : 345312572741

Total Packets FIB Switched : 213554111543

Total Packets ACL Routed : 0

Total Packets Netflow Switched : 346020241

Total Mcast Packets Switched/Routed : 49412512

Total ip packets with TOS changed : 2

Total ip packets with COS changed : 2

Total non ip packets COS changed : 0

Total packets dropped by ACL : 1215

Total packets dropped by Policing : 0

Total packets exceeding CIR : 0

Total packets exceeding PIR : 0

Errors

MAC/IP length inconsistencies : 0

Short IP packets received : 0

IP header checksum errors : 0

TTL failures : 4711

MTU failures : 0

Total packets L3 Switched by all Modules: 559997750297 @ 179965 pps

show plat hard cap cpu :

CPU Resources

CPU utilization: Module 5 seconds 1 minute 5 minutes

3 1% / 0% 0% 0%

4 0% / 0% 0% 0%

5 RP 50% / 49% 46% 45%

5 SP 7% / 0% 5% 5%

Processor memory: Module Bytes: Total Used %Used

3 219980576 53867208 24%

4 219980576 53867448 24%

5 RP 928258080 72475552 8%

5 SP 902888420 144729544 16%

I/O memory: Module Bytes: Total Used %Used

5 RP 67108864 11891816 18%

5 SP 67108864 11891760 18%

I have no ip redirects, no ip unreach, and no ip proxy-arp on these vlan interfaces.

About 90% of the traffic generated is coming in via vlan500 (gige to our bw provider). Most of the traffic than flows to vlan503 (external ip's).

Any suggestions?

berkcom · ‎02-17-2008

I've attached my show int switching as well

berkcom · ‎02-17-2008

Whoops, forgot to mention I am running 12.2(18)SXF7.

berkcom · ‎02-22-2008

Can anyone offer some insight here? Its a bit shocking that a sup720 chokes on less than a gigabit of traffic. I have very little configured on this pair, and I plan to add alot more traffic.

mhurley131 · ‎02-22-2008

Unfortunately more info is going to be needed to lend any real help.

I agree that the CPU usage is a bit on the high side. The usual problems are spanning tree, multicast, ip unreachables, etc...

If possible take a sniffer trace of the incoming traffic to see if you notice anything out of the ordinary. Posting your config may lead to someone noticing something out of the ordinary...

berkcom · ‎02-25-2008

Ok. I am attaching a config of router01. I am going to do a packet sniff of inbound traffic as well.

answanso · ‎02-26-2008

Lets see the following information:

1. show ip traffic (x3 over the course of a min)

2. show proc cpu (while CPU is high)

3. show stack

Lets start with that and we can talk about getting a backplane sniffer capture after I look at the above information.

Email me directly at answanso@cisco.com.

thanks

Anthony

answanso · ‎02-26-2008

Also let me see the following command as well:

1. sh ip nat stat

2. sh ip nat tr verbose

3. show platform hardware pfc mode

4. show mls netflow ip count

Thanks

Anthony