From: Tom Graham (tgraham@xxxxxxxxxxxx)
Date: Mon Feb 05 2001 - 10:57:28 GMT-3
I would be sparing with "clear ip route" ... it can cost you a bunch of time
if you are running RIP/IGRP. I have had RIP labs that with six routers that
took minutes (which seem like hours) to converge. I would think you might
try traceroute (in both directions) first to see if you have multiple routes
between the endpoints (possible with one or more of the alertnate routes
taking you down a black hole).
I have had times in which I thought OSPF was stuck...I just cut the OSPF
config into a notepad file, type "no router OSPF x", and paste the config
back in. (Watch out for loopback/router-ids when doing this though. If you
have added loopbacks or interface IP addresses, your router-id may change).
Often lab books say "save your config and reload the router" -- I am afraid
that would take too much time as well. I have fixed a few problems (maybe)
by using "clear ip ospf r" when there is redistribution going on.
Another suggestion... look at the time stamps on your routes. Link state
routes should be as old as the time up of the router (or routing protocol)
in a stable network (that is, one which doesn't have links going up and down
all the time). Try to pinpoint where the route age is the very
shortest...that is probably close to your problem.
Regards,
TOM GRAHAM
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Chuck Larrieu
Sent: Saturday, February 03, 2001 3:32 PM
To: CCIE_Lab Groupstudy List
Subject: Artifact - when routes appear and disappear - any thoughts?
Have you ever experienced this one:
Complex scenarios, multiple routing protocols, complex redistribution. You
ping from place to place and you are successful. Or maybe partially
successful ( .!!!. for example )
So you troubleshoot. And in the course of troubleshooting, you do some more
tests, and your tests are successful, or sometime unsuccessful. You check
your configurations, your redistribute statements, your route-maps. Things
appear to be what they should be. Test some more. Failure test some more
from different places. Routes are here but not there.
My long winded way of getting around to a phenomenon I am calling
"artifact" - where something seems to be wrong in a routing table or two as
a result of the long history of changes being made. This is particularly
evident, it seems, in protocols such as OSPF and EIGRP, where only changes
are propagated. Something as simple as a clear ip route * on a particular
router ends the problem.
So my question to the group: how do you deal with "artifact"? For those who
have been through the lab, or through ECP1 or ASET, what is your advice
about this kind of concern? What have your mentors suggested as a practice?
Do you make it a habit to issue "clear ip route * " and "clear ip cache"
after making routing related changes, no matter how minor?
Chuck
63 days and counting
A long shot at passing is better than no shot.
Right now that's all I got to get me through,
So I gotta believe!
( paraphrased from Kathy Baille / Baille and the Boys
a song from several years ago )
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:28:37 GMT-3