[c-nsp] RTR/SAA on smaller platforms

[Oliver Boehmer (oboehmer)]

Jee Kay <> wrote on Thursday, August 10, 2006 9:35 PM:

> On 10/08/06, Oliver Boehmer (oboehmer) <oboehmer at cisco.com> wrote:
>> pls try the old "rtr" command syntax, old/new syntax is compared at
>>
http://www.cisco.com/en/US/products/sw/iosswrel/ps5207/products_feature_
>> guide09186a008044d161.html
> 
> .. and suddenly it all makes sense :)
> 
> I'm having a bit of trouble getting through the documentation though.
> Basically what I want to do is ping an IP address every second
> (actually 5 times a sec if possible), timeout around 100ms, and if the
> packet doesn't come back in a timely manner then log or bleet or do
> something interesting that I can somehow detect.
> 
> I'd rather not have to SNMP query every second to see if the last
> packet came back OK if at all possible.... is there any way to store
> say the last minute of traffic and poll that once a minute for dropped
> packets? Or to get it to syslog if such an event occurs? Is 'show rtr
> history' useful at this point?

I don't think you can send that many packets using the echo/ipIcmp
probe. Do you happen to ping to another IOS device which could run the
responder? Then you can take a look at the Jitter probe which has a
configurable interval/num-packets. This probe doesn't support the
history, but you can keep up to 2 hours of stats in the MIB and poll it.
http://www.cisco.com/en/US/products/ps6350/products_configuration_guide_
chapter09186a008044180b.html explains this probe..

	oli