To Redistribute Connected or Not: This is my question

Deadmaster200

Well, going for BSCI on Tuesday.

Just a quick question concerning simulation question.

Let's say you have a simple 3 router scenario and you are told to redistribute between EIGRP and ISIS or whatever. The question does not specifically say to redistribute connected routes. It just has some specifics like the metric for EIGRP and level 1 routes only from ISIS. Should I go ahead and red connected or not? Is it just common sense that you need to red connected for everything to work properly? Or, should I just follow the exact directions of the question?


Thanks

dtlokee

If you do not redistribute the connected routes they will not bee seen in the other protocol because they are seen in the local routing table as connected, not EIGRP or ISIS. If you are already advertising them using a network command you don't need to redistribute them. If you redistribute connected and you use a route-map to supress one or 2 connected interface keep in mind it will also supress not only the connected routes, but the ones originated with the network command. IPv6 redistribution has a "include-conneced" option when doing redistribution to handle the scenario you have laid out here.

EdTheLad

dumping the exam eh!

Deadmaster200

Haha!!!!

Why would you think that? I am going through my lab book and taking practice tests. I was already concerned about this very question when that exact kind of lab popped up in a practice test. The answer was to redistribute connected in one but not the other. Then I thought this practice test just sucks, so I just want to clarify.

DISCLAIMER: I KNOW IT IS NECESSARY FOR THE REAL WORLD. HOW SHOULD I ANSWER FOR THE TEST.


Thanks

tech-airman

Deadmaster200 wrote:

Well, going for BSCI on Tuesday.

Just a quick question concerning simulation question.

Let's say you have a simple 3 router scenario and you are told to redistribute between EIGRP and ISIS or whatever. The question does not specifically say to redistribute connected routes. It just has some specifics like the metric for EIGRP and level 1 routes only from ISIS. Should I go ahead and red connected or not? Is it just common sense that you need to red connected for everything to work properly? Or, should I just follow the exact directions of the question?


Thanks

Deadmaster200,

Since I haven't taken any version of the BSCI exam, I have no comment about the exam therefore also the simulation questions. However, since I'm studying for the 640-861 DESGN exam with the Cisco Press 642-812 BCMSN Self-Study 4th edition book and the CCNP Self-Study: BSCI 2nd edition books, I hope I can provide a useful answer.

According to the Cisco Press CCDA Self-Study: Designing for Cisco Internetwork Solutions (DESGN) book, pages 411~415 covers the topic of "Route Redistribution" and "Route Filtering." So the question to redistrubute or not is the good ol' answer of "it depends." One factor that I think it would depend on is which Enterprise Composite Network Model functional area and modules are being interconnected. So question is, how are the 3 routers in the "simple 3 router scenario" inter-related to each other in relation to the Enterprise Composite Network Model functional area? Is router 1 the Building Access router? Is router 2 the Building Distrubution router? Is router 3 the Campus Backbone router? Are all three routers part of the Campus Backbone? Is one router a Campus Backbone router and the other two Building Distribution routers?

Let's talk about a "simple 1 router" situation with two FastEthernet ports, let's say a Cisco 2621 router. Let's also say that this single router is a Building Distibution router. Let's say the FA0/1 interface leads to a Building Access router running RIP and using IP subnetwork 192.168.1.0/24. Let's say the FA0/0 interface of the Cisco 2621 router is connected to a Campus Backbone router running OSPF and using IP subnetwork 172.16.0.0/16. So let's virtually configure this router.

>enable
Password:
#config t
(config)#int fa0/0
(config-if)#ip address 172.16.0.1 255.255.0.0
(config-if)#no shut
(config-if)#exit
(config)#int fa0/1
(config-if)#ip address 192.168.1.1 255.255.255.0
(config-if)#no shut
(config-if)#exit
(config)#router rip
(config-router)#network 192.168.1.0
(config-router)#exit
(config)#exit
#

Now, the abovementioned configuration steps are at the current CCNA level. Also at the CCNA level, you should know that due to a missing "(config-router)#network 172.16.0.0" statement under the "(config)#router rip" section, there's going to be routing protocol connectivity issue because the RIP routing protocol doesn't know about network 172.16.0.0, even though it's directly connected. So you have to add the network 172.16.0.0 explicitly under the "(config)#router rip" section as follows.

#config t
(config)#router rip
(config-router)#network 172.16.0.0
(config-router)#exit
(config)#exit
#

Now, once again at the CCNA level, you should now know that packets sent from the 192.168.1.0 network to the 172.16.0.0 network should be dynamically routed by RIP, and vice versa. In essence, the RIP routing protocol for the 192.168.1.0 network is "redistributed into" the RIP routing protocol for the 172.16.0.0 network and vice versa.

So what does this have to do with redistributing between two routing protocols? Well, for starters, remember that for some routing protocols, either the router interface belongs to the routing protocol or the whole router belongs to the routing protocol. For example, router interfaces belong to the RIP, IGRP, and OSPF routing protocols. I believe the whole router belongs to the ISIS routing protcol. So let's reconfigure the example virtual router for OSPF on the Fa0/0 interface.

#config t
(config)#router rip
(config-router)#no network 172.16.0.0
(config-router)#exit
(config)#router ospf 2
(config-router)#network 172.16.0.0 0.0.255.255 area 0
(config-router)#exit
(config)#exit
#

This configuration leads to a loss of connectivity situation because even though both fa0/0 and fa0/1 are configured with valid IP addresses and the directly connected networks are known by two separate routing protocols, but because the routing protocols weren't talking to each other led to the loss of connectivity issue that developed. Just like the abovementioned RIP example with only one interface's network configured under the "(config)#router rip" statement. I was messing around with my Cisco 2621 router with a similar configuration as above which is why I know about the "loss of connectivity" issue. :P Since at the time I didn't know how to configure redistribution, I reverted back to the single routing protocol configuration and my loss of connectivity issue went away which is how I'm typing this reply and posts since I restored connectivity.

So in summary, some factors to consider to decide about route redistribution are which Enterprise Campus Network Model functional areas and modules are involved and which routing protocols are involved. By specifically NOT configuring route redistribution may be a form of route filtering.

I hope this helps.

dtlokee

tech-airman:

I am not sure if his question was how to configure a router but more along the lines of if you have 3 routers (a 1 router scenario with routing protocols is not very practical) and you ONLY want to run OSPF between R1 and R2 and you ONLY want EiGRP btween R2 and R3.

 ------------- R1 ----------------- R2 ---------------- R3 ----------------
10.1.1.0/24         10.1.2.0/24           10.1.3./24          10.1.4.0/24

now of course we can run a single protocol to accomplish our goal, but that is not what we are doing here, we are going to run ospf between R1 and R2 and EIGRP between R2 and R3

Hostname R1
!
Interface eth0/0
  ip address 10.1.1.1 255.255.255.0
!
Interface eth0/1
 ip address 10.1.2.1 255.255.255.0
!
router ospf 1
 network 10.1.1.1 0.0.0.0 area 0
 network 10.1.2.1 0.0.0.0 area 0

===========

Hostname R2
!
interface eth0/0
 ip address 10.1.2.2 255.255.255.0
!
interface eth0/1
 ip address 10.1.3.1 255.255.255.0
!
router ospf 1
 network 10.1.2.2 0.0.0.0 area 0
!
router eigrp 100
 network 10.1.3.1 0.0.0.0 area 0

===============

Hostname R3
!
interface eth0/0
 ip address 10.1.3.2 255.255.255.0
!
interface eth0/1
 ip address 10.1.4.1 255.255.255.0
!
router eigrp 100
 network 10.1.3.2 0.0.0.0
 network 10.1.4.1 0.0.0.0

now at this point we'll have routes in the routing table of R1, R2 and R3 learned from the other routers. However R1 will not have knowledge of 10.1.3.0/24 and 10.1.4.0/24, time to redistribute.

R2
=========

router ospf 1
 redistribute eigrp 100 subnets
!
router eigrp 100
 redistribute ospf 1 metric 1500 1500 255 1 1500

now if for some reason you're using a route map to filter what routes are being learned or redistributed you want to ensure the connected routes are included or you can end up removing them from the redistributed routes.

For the question of should you or should you not redistribute on the exam, I'll pass on some wisdom a lab proctor gave me, "do what you think is right."

Good luck on your exam!

mikej412

If we were talking about a CCIE Practice Lab, I'd go for "universal connectivity and world peace."

But for the BSCI exam -- I'd hope that the instructions were more clear on what they wanted to accomplish. But I do think it would be like a CCIE Lab -- you don't get penalized for over configuration as long as you don't violate a requirement.

Deadmaster200

Haha!!!!

Guys, guys. Thanks for the long posts. But that is not what I am asking here. I am not unclear on the technology and concepts. I just want to clarify what Cisco is expecting for this specific kind of question.

The sim questions are just an exercise to test a specific simple and basic cookie-cutter configuration (at least that is my experience so far, other, more difficult tests might have complex sim questions, who knows?). So, I would think it prudent to know what they are expecting for the specific cookie-cutter config.

Thanks Mike!!!! That is exactly what I needed. To 'red con' it is!!!