Re: IP cache issue

felipe.oliveira · ‎01-30-2007

My two 6500 run in redundancy, with vlans balanced between them. One cat has the following cpu load:

RJ6509_1#sh proc cpu

CPU utilization for five seconds: 36%/32%; one minute: 47%; five minutes: 42%

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

1 4 10 400 0.00% 0.00% 0.00% 0 Chunk Manager

2 1888 32256 58 0.00% 0.00% 0.00% 0 Load Meter

3 152 517 294 0.00% 0.00% 0.00% 0 Constellation Ba

4 652700 42788 15254 0.00% 0.29% 0.24% 0 Check heaps

5 4 39 102 0.00% 0.00% 0.00% 0 Pool Manager

While the other:

RJ6509_2#sh proc cpu

CPU utilization for five seconds: 17%/14%; one minute: 17%; five minutes: 19%

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

1 0 12 0 0.00% 0.00% 0.00% 0 Chunk Manager

2 2152 55274 38 0.00% 0.00% 0.00% 0 Load Meter

3 32 513 62 0.00% 0.00% 0.00% 0 Constellation Ba

4 692184 54482 12704 0.24% 0.21% 0.23% 0 Check heaps

5 16 62 258 0.00% 0.00% 0.00% 0 Pool Manager

6 0 2 0 0.00% 0.00% 0.00% 0 Timers

When I issue "sh ip cache" on the primer, that is what I see:

RJ6509_1#sh ip cache

IP routing cache 3 entries, 528 bytes

2994 adds, 2991 invalidates, 0 refcounts

Minimum invalidation interval 2 seconds, maximum interval 5 seconds,

quiet interval 3 seconds, threshold 0 requests

Invalidation rate 0 in last second, 0 in last 3 seconds

Last full cache invalidation occurred 21:43:47 ago

Prefix/Length Age Interface Next Hop

200.150.148.80/30 01:00:40 Vlan14 200.150.148.76

200.179.200.0/24 00:05:33 Port-channel3 200.155.20.90

201.36.206.0/24 00:04:41 Port-channel3 200.155.20.90

While the second one:

RJ6509_2#sh ip cache

IP routing cache 131 entries, 23948 bytes

13777 adds, 13646 invalidates, 0 refcounts

Minimum invalidation interval 2 seconds, maximum interval 5 seconds,

quiet interval 3 seconds, threshold 0 requests

Invalidation rate 0 in last second, 0 in last 3 seconds

Last full cache invalidation occurred 3d04h ago

Note that the first with higher cpu usage, only has 3 entries, while the other has 131 entries.

The configuration on both is the same. Can it be a hardware problem ?? What can it be ??

lgijssel · ‎01-30-2007

As a first step, you should check the complete output from "sh proc cpu" to see which process is actually taking the most cpu. This will give a first impression of the direction to search in.

Regards,

Leo

felipe.oliveira · ‎01-30-2007

I've already done this. All the process are normal. At least, I guess no process justify the high utilization:

Thanks

lgijssel · ‎01-31-2007

You are correct, nothing there. The link below might suggest a possiblke cause:

http://www.cisco.com/en/US/products/hw/switches/ps708/products_tech_note09186a00804916e0.shtml

Regards,

Leo

felipe.oliveira · ‎02-02-2007

I guess I'm having a loop problem in the network. I just put a certain vlan in shut and the proc cpu went down imediatly.

Thank you