From: Edward Moss (teklnk@xxxxxxxx)
Date: Fri Apr 06 2001 - 22:35:39 GMT-3
I haven't taken the lab yet, so I assume that I cant break the NDA.... :o)
My advise... Identify what is broken then work the physical layer up.
In the real world, for networks I support it can be "by the seat of my
pants"... but in customer's networks here is wht I do.
1. Get the documentation (helps if interfaces and port numbers are
documented). I believe it is critical to know what the network was designed
to do.
2. Get a copy of each config for a baseline (is the startup and running
config the same?). Even if it is bad... you can go back to it if necessary.
3. On each device, check the interface status, ensure UP/UP - DCE vs. DTE,
clock rate, keepalives?
4. Look at CDP if possible, check IP between devices. Verify the configs
(IP/prefix) against the docs.
5. Check the routing protocols... what is running on each device, what
should be advertised, is it in the appropriate route table.
6 If necessary, remove route redistrubitions, access lists, etc... If
there are access lists... know what each one is doing.
Of course, each protocol and routing protocol have their own querks.
Above all... document every change, why it was changed and note the results.
Did the change have the desired effect? If not, change it back (or have a
good reason for keeping it in.)
I think the Cisco Press book "Internetwork Troubleshooting Handbook" is a
great on-the-job book. It details common problems as well as the physical
layer up approach.
Ed
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:29:42 GMT-3