This is post 4 so far in this lab. Check out the first three posts (post 1, post 2, and post 3) for the proper background. Today’s post: some analysis of major task 1, in which you predict why the routing tables on R1, R2, and R3 look like they were shown to be in lab post 3. Details below the fold.
A Suggested Strategy: Start with Connected Routes
The biggest variable in figuring out why a router lists some routes, and not others, is what the routing protocol. Lots of events effect the routing protocol: its configuration on the local router, on the remote routers, the data link configurations on the various router interfaces, current interface status, even access control lists (ACLs). But logic a router uses to add connected routes uses less steps, less logic, and leads you to fewer variables to investigate. So I’ll suggest that you do a quick survey of the connected routes.
The other great troubleshooting benefit of knowing the connected routes is that whatever the routing protocols learn, their routes are always worse than connected routes. IOS considers connected routes to have an administrative distance (AD) of 0, meaning connected routes are the most believable routes. IOS defaults its AD for RIP to 120, so any RIP-learned routes will be worse that any connected routes. So if a connected route is missing from a router’s routing table, the root cause is *not* the routing protocol itself. In other words, the RIP-2 configuration, nor the RIP-2 messages, can cause a router to not list a connected route.
Specifically, I’d suggest looking at the network diagram, and predicting the connected routes, and compare those to the routing tables.
Strategy Step 2: Assume a Config, Chase the Route Advertisements, and Look for Surprises
As a second step, then start at one router, and predict what routes it would advertise to each neighbor when using default configuration. With RIP-2, we already know what is configured from lab post 2, and we know from lab post 3 that the config has changed. At this point, you’re looking for the differences. What should each router advertise, and each router learn, using that known configuration? How is that different from what they did learn? And finally, what configuration change could have caused those changes?
Step 1 Analysis: Missing a Connected route on R1 and R2
Let’s walk through this step for this lab. As a reminder, Figure 2 repeats the topology, listing the subnet IDs of the subnets for easier reference.
Figure 2: CCENT Lab 3 Topology
The lab exercise has already told us that no other interfaces exist beyond the figure. All three routers have three interfaces, so all the routers should have three connected routes. However, R1 and R2 only have two connected routes; both are missing their route for subnet 10.9.6.0/25, which is the subnet between R1 and R2 on a serial link.
Unfortunately, the output of the show ip route command does not tell you why the connected route is missing. If this were part of a Sim or Simlet question, how would you proceed? I’ll close this post with a few comments, and leave it to you as an exercise.
Requirements for a Connected Route
For a router to list a connected route in its routing table, the following must be true:
- An interface must have an IP address/mask configured
- The interface must be in an up/up state
Of these two criteria, the first is much easier to investigate. A quick show interfaces or show ip interfaces brief command tells you if the interface has an IP address configured. Those same commands will also tell you whether the interface state is up/up or not. However, if the interface is not up/up, finding the root cause requires more investigation, and for the purposes of this lab, I’m going to stop that part of the investigation here.
So, back to the main point of lab task 1: what changed in the configuration that caused the routing tables shown in post 3, for lab task 1 (Examples 7, 8, and 9)? You tell me which of the following could have resulted in the missing connected routes on R1 and R2:
- R1 (only) being configured with a no ip address command under interface s0/0/0
- R2 (only) being configured with a no ip address command under interface s0/0/1
- R1 and R2 being configured with a no ip address command under their respective interfaces
- R1 (only) being configured with a shutdown command under interface s0/0/0
- R2 (only) being configured with a shutdown command under interface s0/0/1
- R1 and R2 being configured with a shutdown command under their respective interfaces
- Other stuff (you tell me; there are others)
I’ll watch for any other theories you have!