Traceroute latency values for traffic hitting 4506 SupIV SVI's

Unanswered Question
Aug 2nd, 2010

We have noticed in that icmp echo-replies and traceroutes that hit the real ip (rip) of the SVI on our 4506 supervisor have unusually high response times, but traces and pings through the switch respond with normal/expected latency values.

It is my understanding that the Supervisor IV engine places a lower value on processing packets to an svi than it does through the svi, but I need some documentation confirming this to appease our delvelopers, or any other solutions that may eliminate this delay.

Appreciate in advance any relevant responses.



I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Jayakrishna Mada Mon, 08/02/2010 - 14:49

Hi Scott,

The design of the switch processes is such a way that "ICMP" gets low priority when compared with other proceses. So ping to the switch itself is not a good idea. Unfortunately there is no document that backs this. There are lot of threads on internet which talk about this, you can find a good explanation on this previous post.

Not all traffic hitting the SVI will have latency issue.



This Discussion