Answers: RIPv2 Options 1
Know how to limit the number of RIP routes to each destination? How to enable RIP on an interface – but to not send Update messages out that interface? How to make RIP not summarize routes for classful networks? The lab asked you for all three; check here to see if your answered matched ours.
Answers
Figure 1: Topology for this Lab
Example 1: R1 Config
|
1 2 3 4 5 6 7 8 9 10 11 |
router rip version 2 network 10.0.0.0 network 20.0.0.0 network 192.168.1.0 network 192.168.2.0 no auto-summary passive-interface GigabitEthernet0/2.1 passive-interface GigabitEthernet0/2.2 passive-interface GigabitEthernet0/2.3 maximum-paths 1 |
Example 2: R2 Config
|
1 2 3 4 |
router rip version 2 network 20.0.0.0 maximum-paths 1 |
Example 3: R3 Config
|
1 2 3 4 5 6 7 |
router rip version 2 network 10.0.0.0 network 20.0.0.0 no auto-summary passive-interface GigabitEthernet0/2 maximum-paths 1 |
Commentary
There are a number of different options that can be configured for a dynamic routing protocol, some of these including limiting the maximum number of paths that can be inserted into the routing table for the same destination, limiting the interfaces out which the routing protocol will send routing protocol advertisements, as well as limiting whether supporting routing protocols will automatically attempt to summarize IP networks at classful boundaries. This lab asked you to consider all three, as well as enabling the core functions of RIPv2 on three routers.
To enable RIPv2, ignoring all these options for now, you need some familiar commands. The requirements asked for RIP, and specifically that you use RIP Version 2 (RIPv2) only. As a result, all three routers need the same two commands:
- router rip
- version 2
Additionally, each router needed a combination of network commands: one for each class A, B, or C network to which that router was connected. R1 has three subinterfaces, each with an IP address in a different classful network (192.168.1.0, 192.168.2.0, and 10.0.0.0), so R1 needs a network command for each. R1 also has an interface with an IP address in class A network 20.0.0.0, requiring a network 20.0.0.0 command.
Similarly, on R2 and R3, R2 has interfaces with addresses but both addresses are in class A network 20.0.0.0, requiring only a single network 20.0.0.0 command. R3’s interfaces have addresses in two different networks, requiring two commands: network 10.0.0.0 and network 20.0.0.0.
So far, the discussion has focused on the core RIPv2 configuration. Next, the discussion turns to the optional features.
Auto-summarization occurs on routers that have multiple interfaces with RIP enabled and the interfaces are in different classful networks. That fact is true of both R1 (connected to four different classful networks) and R3 (connected to two different classful networks). In fact, this topology and addressing plan would have problems if you left the default of auto-summary enable. Because R1 has only a single interface in the 10.0.0.0/8 network and there is another 10.0.0.0/8 network being advertised on R3, with the default of auto-summary, R2 would learn two routes for 10.0.0.0/8: one from R1, and one from R3. Clearly, only a subset of 10.0.0.0/8 is off R1, and a different subset exists of R3.
To disable RIP automatic summarization use the no auto-summary command. Note that this command is needed only on R1 and R3. R2, which does not connect to multiple classful networks, would not be affected by the command.
The next requirement specifies that all interfaces and subinterfaces that are not connecting to another router should not send RIP updates. R1 has three different subinterfaces that are configured with IP addresses that need to be disabled, and R3 has one interface. To disable RIP multicasts on these subinterfaces use the passive-interface GigabitEthernet0/2.1, passive-interface GigabitEthernet0/2.2, and the passive-interface GigabitEthernet0/2.3 commands. On R3, use the passive-interface GigabitEthernet0/2 command.
The last requirement was to limit the maximum number of paths that will be used to the maximum that will be required for the network displayed in the figure. That particular requirement may have been either too tricky or too easy. With the topology in the figure, only one route could possibly exist for each destination, because there are no redundant paths in the network. This requirement was mainly to get you thinking about the maximum-paths RIP subcommand.
Given that only one equal-metric route to each destination could exist in this topology, per this requirement, you should have configured the maximum-paths 1 command on each router.
Why R2 does not need the no auto-summary command?
Since it is connected to two different classful subnets (20.10… and 20.20…)
Hi Leo,
The auto-summary command has an effect only in cases where the router has interfaces connected to different IP networks. R2 is connected to two interfaces, both of which are connected in the same class A network 20.0.0.0. In contrast, both R1 and R3 have interfaces in more than 1 class A, B, or C network, so the auto-summary setting has an impact.
Wendell
Hello!
In my answer for R1, instead of disabling RIP updates on all 3 of the sub-interfaces with:
passive-interface g0/2.1
passive-interface g0/2.2
passive-interface g0/2.3
I disabled it only on physical interface with:
passive-interface g0/2
Is this a viable answer? Tks!
Hi Wendell,
passive-interface default
no passive-interface g0/1
is it correct on R1? instead of disabling RIP updates on all 3 of the sub-interfaces,
Many thanks
I have three Routers: R1,R2 and R3. They are connected through their fa0/0 (R1 to R2) and
fa0/1 (R2 to R3) interfaces.
R1 has connections to the following networks through these interfaces:
fa0/0: 192.168.1.254/24
fa0/1: 10.255.255.254/8
fa0/2: 172.16.1.254/16
fa0/3: 172.17.1.254/16
for R2:
fa0/0: 192.168.1.253/24
fa0/1: 192.168.2.254/24
fa0/2: 192.168.3.254/24
fa0/3: 172.18.1.254/16
and for R3:
fa0/1: 192.168.2.253/24
fa0/2: 192.168.4.254/24
fa0/3: 172.19.1.254/16
I enabled RIPv2 on both routers.
Each one of them fulfil the criteria of “Directly connect to more than one classful network”
Would “no auto-summary” command really have an impact on them?
———————————————————————————————–
On the CCENT/CCNA ICND1 100-105 book, page 466, was established that:
1.- The no auto-summary command has an impact only on routers that directly connect
to more than one classful network.
2.- However, the command is needed only if a discontiguous classful network exists
However, I would make some remarks:
For the first point I’d rather say (according with the example I’ve written down above)
“The no auto-summary command has an impact only on routers that directly connect
to more than one classful network WHICH HAVE BEEN SUBNETED”
For the second one, on the same book, page 465, there’s an example which
mentions a case without having a discontiguous network, but “no auto-summary” is needed so a router
which has AT LEAST ONE subnet of a class network at a boundary of this one, on the router, can
advertise subnet and its mask, and to do not allow the router advertising the entire classfull network
On the book, not issuing “no auto-summary” in this case is considered a mistake (although the
routing works properly).
So, wouldn’t it be better expressing those requirements just as:
“‘No auto-summary’ command must be used on a router connected to some others which are part
of the same RIP routing process and connects at least two different classful networks, which
at least one of them has been subneted and would be announced by RIP”
or
“‘No auto-summary’ command must be used on a router connected to some others which are part
of the same RIP routing process and is the boundary among at least two different classful networks which
at least one of them has been subneted”
?
By the way, on that page (465) of the mentioned book, there’s an explanation about auto-summary
issues. It uses figure 19-18 on 462 page. All routers on that figure have gigabitethernet interfaces,
On page 465 appear, at Example 19-18, two “show ip route” displays, which were issued
on one of the routers of figure 19-18. However, routes of those displays inform that where learned
through non existence serial interfaces.
I think that it is worth checking out that matter.
Thanks on advance
Hi Hector,
Wow, that’s a long post. Let me try and cull your questions/statements out and make a reply.
Your comment on your #1: I agree. I do think the book makes that same point, just not in the statements from the book that you repeated here. The chapter does point out that auto-summary only matters if subnets of the network exist.
Your comment on your #2: Again, I think the context does exist in the book, just not where you mentioned the text. The book said the no auto-summary is only needed if there are discontiguous subnets. That’s still true. But that’s the case of when it is “needed”. The example you then reference on page 465 is not about the need for the feature. I think it’s clear in the text that the section is about the mechanism, and a common mistake with the mechanism of thinking that it applies to routers that are not connected to multiple networks. I don’t see anything in the section that sets up the example as one that’s about whether the feature is needed aka useful. (That topic had been covered already.)
On your longer summary and suggested wording, I would agree with your wording. If my goal was to summarize all the ideas in one complete statement, I’d be good with both options. Whether those are better or worse is of course a matter of intent and goal of what I’d like to communicate in that section. I can certainly add it to the list of things to consider for next edition – that’s a list I start pretty much the week that I wrap the books. 🙂
Finally on the mistake you noted in example 19-18, about it listing serial interfaces instead of Gigabit, I agree. I had come across that one before as it turns out, but I’m always happy to hear of those so I can clean those up. FYI, you can find those (and some others) in the errata file that’s posted on the book’s web page at ciscopress.com.
Thanks for the post,
Wendell
Hello Mr. Odom,
Dear respects for your work. I am preparing the CCNA certification and using it all time.
I will try to be clear and consice: I am using Packet Tracer tool to reproduce this lab. Through my results if I do not disable auto-summary on R2: I find in R1 routing table a route for 10.0.0.0/8 via 20.10.20.2, ans in R3 routing table a route for 10.0.0.0/8 via 20.20.30.1.
In addition, if I issue “debug ip rip” on R1 & R3, the Ripv2 sending updates to respectives G0/1 interfaces show both directly connected network 10.10.10.0/24 & 10.30.30.0/24 included, which seems to be convenient.
Furthermore, I conluded R2 would be the root cause, indeed if I issue “debug ip rip” on R2, I see now the RIPv2 sending updates going toward Gi 0/1 & Gi 0/2 include the network 10.0.0.0/8 in both cases, not my subnets 10.10.10.0/24 & 10.10.30.0/24.
However, through “ip rip database” and “show ip route” issued on R2 I find the networks 10.10.10.0/24 & 10.30.30.0/24, and not a 10.0.0.0/8.
Those elements, right now, make me believe in my configuration that R2 summarizes the networks to be sent (not received) to outgoing interfaces.
I’ve checked my configuration versus the lab’s rules and results, no difference of configuration on my point of view.
If I turn-off auto-summarization on R2, 10.0.0.0/8 disappear from RIPv2 updates sent by R2 and I can find my 10.10.10.0/24 & 10.30.30.0/24 networks. Compliant routes now appear in R1 & R3 routing tables.
Hoping that I did not carried a mistake in the configuration leading to this result, I don’t like the feeling to maybe wasting your time.
But with the goal to at worst understand my mistake or the phenomenon, would you accept sharing your opinion about this result ?
Thank you very much.
Best regards
Hi QLacroix, a similar “problem” was discussed in the exercise Ripv2 Basics 1 (https://blog.certskills.com/ccent/cl122-answer/).
It seems that, if a router has auto-summary enabled, it’s going to advertise summary routes of its directly connected routes as long as their interfaces belong to different subnets of classful networks, and also, it will advertise summary routes of remote subnets learned through RIP despite these arrived as subnets from routers where auto-summary is enabled.