Traceroute and load balencing

From: alsontra@hotmail.com
Date: Thu Jan 29 2004 - 23:41:18 GMT-3


Group,

I think I know the answer to this one, but I would like to be sure. The
example topology is as follows:

                  ________
Host A ----R1_______ R2------ Host B

R1 and R2 are doing load balancing OSPF, EIGRP, whatever its not important.

Host A pings Host B and the trace that comes back should not be exactly linear
correct? Load balancing would cause some packets to take route 1 vs. route 2.
(round robin)

I have a similar situation in my home lab and want to make sure what I'm
seeing is actual the result of load balancing.

The traces look something like this when I have load balancing enabled.

Tracing the route to 111.1.1.1

  1 160.1.1.6 4 msec
    150.1.1.2 0 msec
    160.1.1.6 0 msec
  2 141.1.1.5 32 msec
    141.1.1.9 36 msec
    141.1.1.5 36 msec
  3 171.1.1.3 36 msec 36 msec 32 msec
  4 180.1.1.8 36 msec

And when only one link is available the trace looks like this:

Tracing the route to 111.1.1.1

  1 160.1.1.6 0 msec 4 msec 4 msec
  2 141.1.1.9 28 msec 32 msec 36 msec
  3 171.1.1.3 32 msec 36 msec 32 msec
  4 180.1.1.8 36 msec * 32 msec

All ping attempts are successful and no data is ever lost. I'd just like to
verify that traceroute is actually displaying load balancing.

Alsontra



This archive was generated by hypermail 2.1.4 : Mon Feb 02 2004 - 09:07:52 GMT-3