Classless and Classful routing headachs + super confused

datchcha

CCNA Study Guide
By Todd Lammle
Edition: 4

Ok, late last night I was doing some last minute studies, because I am preparing to take the 801 on Saturday, but I read over some items about RIP v1 & v2, IGRP and EIGRP that I must have missed in the past. I am super confused, and I am thinking it is a printing mistake, but I am unsure. I would like to run this by everyone and see what they have to say.

Model 1:
R1 – S0/0: 10.0.1.1/24, fa0/0: 10.0.2.1/24. R2 – S0/0 10.0.1.2/24, fa0/0 192.168.1.1/24

RIP: I already understand that RIP only supports classful networks. All subnet masks have to match. But look at pg. 252, “this is because RIP version 1 doesn’t send update with subnet mask information in tow.” Pg. 253 “You would only type in the classful network address of 172.16.0.0 and let RIP find the subnets and place them in the routing table.”

Ok, If RIP doesn’t send any subnet information, how is it going to find the subnets and place them in the routing table?

Pg. 258 “RIPv2, unlike RIPv1, is a classless routing protocol (even though it is configured as classful, like RIPv1), which means that is sends subnet mask information along with the route update.”

So with RIP configuration I am really confused now. If I use the above model, would I configure the networks as the following:

Router rip
Network 10.0.0.0
Network 192.0.0.0

Pg. 260 “(side note) You absolutely must remember that you type a classful network number in when configuring IGRP.”

Pg. 261 “IGRP uses classful routing, which means that subnet mask information isn’t sent along with the routing protocol updates.

Pg. 261 “(side note) If you’re using 172.16.0.0/24 network, know that if you did type in the subnet 172.16.10.0, the router would accept it and then change the configuration to a classful entry of 172.16.0.0. But don’t do that – at least no on the exam! It’s definitely not so forgiving and will simply mark your answer wrong if you type the wrong network number. I cannot stress this enough: think classful if you are suing subnets!”

Configuring EIGRP
Pg. 298 “Remember – as with IGRP, you use the classful network address, which is all subnet and host bits turned off.”

model 2:
R1 - S0/0: 10.0.0.1/30, fa0/0: 192.168.0.1/24. R2 - s0/0:10.0.0.2/30, fa0/0: 172.16.0.1/24

I would have to use EIGRP or OSPF for this model. I would not have to worry about auto summary, since both distance networks are different right? But what would the commands be to configure this model along with the first model.

Thank you.

Pash

Ok, If RIP doesn’t send any subnet information, how is it going to find the subnets and place them in the routing table?

Auto summary, its on by default on all cisco devices, it picks the mask for you using the fundamentals of IP addressing. With Ripv1 you cant turn it off.

Look at the cisco docs and your notes and study guides for how auto summarization works if you are unclear fella.

I would have to use EIGRP or OSPF for this model. I would not have to worry about auto summary, since both distance networks are different right? But what would the commands be to configure this model along with the first model.


As the two private networks are completely different auto summarzation doesnt really play any part here no, correct. If you started using VLSM and had multiple private networks with the same summarized bits then yes it would

As from the pasted passages from the book, they all look fine to me my friend.

Good luck with your exam.

datchcha

Pash wrote:

Ok, If RIP doesn’t send any subnet information, how is it going to find the subnets and place them in the routing table?

Auto summary, its on by default on all cisco devices, it picks the mask for you using the fundamentals of IP addressing. With Ripv1 you cant turn it off.

Look at the cisco docs and your notes and study guides for how auto summarization works if you are unclear fella.

I would have to use EIGRP or OSPF for this model. I would not have to worry about auto summary, since both distance networks are different right? But what would the commands be to configure this model along with the first model.


As the two private networks are completely different auto summarzation doesnt really play any part here no, correct. If you started using VLSM and had multiple private networks with the same summarized bits then yes it would

As from the pasted passages from the book, they all look fine to me my friend.

Good luck with your exam.

so with RIP, if my network is 10.0.1.1, and 192.168.20.1, i would configure it as followed:

router rip
network 10.0.0.0
netowrk 192.16.0.0

When would i configure it like the the following:

network 10.0.1.0
network 192.168.20.0

Pash

datchcha wrote:

Pash wrote:

Ok, If RIP doesn’t send any subnet information, how is it going to find the subnets and place them in the routing table?

Auto summary, its on by default on all cisco devices, it picks the mask for you using the fundamentals of IP addressing. With Ripv1 you cant turn it off.

Look at the cisco docs and your notes and study guides for how auto summarization works if you are unclear fella.

I would have to use EIGRP or OSPF for this model. I would not have to worry about auto summary, since both distance networks are different right? But what would the commands be to configure this model along with the first model.


As the two private networks are completely different auto summarzation doesnt really play any part here no, correct. If you started using VLSM and had multiple private networks with the same summarized bits then yes it would

As from the pasted passages from the book, they all look fine to me my friend.

Good luck with your exam.

so with RIP, if my network is 10.0.1.1, and 192.168.20.1, i would configure it as followed:

router rip
network 10.0.0.0
netowrk 192.16.0.0

When would i configure it like the the following:

network 10.0.1.0
network 192.168.20.0

Both would work when configuring RIP, the second option would just be summarized, which you understand. So when would you configure it like the second option......never really...in good practice it should be like the first....at least for the exam.

datchcha

Pash wrote:

datchcha wrote:

Pash wrote:

Ok, If RIP doesn’t send any subnet information, how is it going to find the subnets and place them in the routing table?

Auto summary, its on by default on all cisco devices, it picks the mask for you using the fundamentals of IP addressing. With Ripv1 you cant turn it off.

Look at the cisco docs and your notes and study guides for how auto summarization works if you are unclear fella.

I would have to use EIGRP or OSPF for this model. I would not have to worry about auto summary, since both distance networks are different right? But what would the commands be to configure this model along with the first model.


As the two private networks are completely different auto summarzation doesnt really play any part here no, correct. If you started using VLSM and had multiple private networks with the same summarized bits then yes it would

As from the pasted passages from the book, they all look fine to me my friend.

Good luck with your exam.

so with RIP, if my network is 10.0.1.1, and 192.168.20.1, i would configure it as followed:

router rip
network 10.0.0.0
netowrk 192.16.0.0

When would i configure it like the the following:

network 10.0.1.0
network 192.168.20.0

Both would work when configuring RIP, the second option would just be summarized, which you understand. So when would you configure it like the second option......never really...in good practice it should be like the first....at least for the exam.

Could i get away with configuring it like the following since auto summary is enabled by default?

router rip
network 10.0.0.0
network 192.0.0.0

Would it be currect to say the following:

router egirp 100
network 10.0.1.0
network 192.168.20.0
no auto-summary

I am just a little confused because in the Todd Lammle book, he is showing the following when configuring EIGRP:

router eigrp 10
network 192.168.20.0
netowrk 192.168.30.0
network 192.168.40.0

I truly appreciate the help...Cheers!!!

datchcha

Now I am lost.

I fired up my routers, and i tried to configure the following on a 2650 and this is what i got.

R1 - S0/0:10.0.1.1/24, fa0/1:192.168.10.1/24

I configured the interface in the following ways

router rip
network 10.0.0.0
network 192.0.0.0

the 10.0.0.0 worked fined and sent up dates out, was able to confirm this by using debug ip rip, but 192.0.0.0 was not sending up dates. I then tried

no network 192.0.0.0
network 192.168.0.0

Updates were not sent until i changes the configuration to

no network 192.160.0.0
network 192.168.10.0

then updates started to show on debug ip rip.

LOL...I am so confused at this point.

dtlokee

I think the first part of your confusion comes from what the network command does. It tells the router what interfaces should participate in a routing protocol, not what networks (or subnets) to advertise. The router will advertise the network (or subnet) associated with the interface.

So let's say you have:

E0 - 192.168.100.32/27
E1 - 192.168.100.64/27
S0 - 192.168.100.96/27

Entering the "network 192.168.100.0" command will tell the router to run the routing protocol on any interfaces that match the 192.168.100.0 network or any subnet of the 192.168.100.0. So what will the router advertise? All 3 subnets will be advertised to the neighbors.

The idea of classful routing protocols would indicate the protocol does not carry the mask in the updates so the receiving router will assume it has the same mask as the interface that the update was received on. This also expalains the behavior of a classful protocol when a router wants to send an advertisement for one classful network out an interface with a different classful network address, it will summarize to the classful boundry (no subnets)

datchcha

dtlokee wrote:

I think the first part of your confusion comes from what the network command does. It tells the router what interfaces should participate in a routing protocol, not what networks (or subnets) to advertise. The router will advertise the network (or subnet) associated with the interface.

So let's say you have:

E0 - 192.168.100.32/27
E1 - 192.168.100.64/27
S0 - 192.168.100.96/27

Entering the "network 192.168.100.0" command will tell the router to run the routing protocol on any interfaces that match the 192.168.100.0 network or any subnet of the 192.168.100.0. So what will the router advertise? All 3 subnets will be advertised to the neighbors.

The idea of classful routing protocols would indicate the protocol does not carry the mask in the updates so the receiving router will assume it has the same mask as the interface that the update was received on. This also expalains the behavior of a classful protocol when a router wants to send an advertisement for one classful network out an interface with a different classful network address, it will summarize to the classful boundry (no subnets)

Ok, so a classful address would be 192.168.100.0, while a classless address would be 192.168.0.0 or 192.0.0.0?

Thanks.

dtlokee

Classful would indicate that you are using the default or natural mask and the address is truncated to the classful boundry. If you had a 192.168.100.32/27, the classful address would be 192.168.100.0/24. Anytime you are using a mask that is shorter than the default mask, it is classless at that point (CIDR). This would be used to create a block of networks (sometimes called a supernet) that could be allocated to a company.

For example, let's say we're a start up ISP, and we want some addresses from ARIN. We need a total of 6000 addresses but ARIN only has class 'C' addresses for us. Each class C has 254 hosts addresses so they can give us a block of 32 class C networks (it needs to be an even multiple of 2**n). The resulting mask would be a /19 which is shorter than the natural mask of /24. This allocation could be somthing like 207.44.32.0/19, if you do the math this creates 1 large network starting at 207.44.32.0 ending at 207.44.63.255, a total of 8190 addresses.

gabrielbtoledo

datchcha, you got confused by the class A and C.

Now I am lost.

I fired up my routers, and i tried to configure the following on a 2650 and this is what i got.

R1 - S0/0:10.0.1.1/24, fa0/1:192.168.10.1/24

I configured the interface in the following ways

router rip
network 10.0.0.0
network 192.0.0.0

the 10.0.0.0 worked fined and sent up dates out, was able to confirm this by using debug ip rip, but 192.0.0.0 was not sending up dates. I then tried

no network 192.0.0.0
network 192.168.0.0

Updates were not sent until i changes the configuration to

no network 192.160.0.0
network 192.168.10.0

then updates started to show on debug ip rip.

LOL...I am so confused at this point.

Class A subnet mask 255.0.0.0
So your address 10.0.1.1/24 = 10.0.0.0

Class C subnet mask 255.255.255.0
So your address 192.168.10.1/24 = 192.168.10.0

Hope that helps.

gojericho0

Here's a couple more good links to help you reinforce how classful routing works. Like dtlokee stated earlier remember what the network command does and does not do. It does not advertise the IP address you add after the network command, it just uses that routing protocol on any interfaces that fall under that network. That took me longest time to understand as well and I think is a common misconception for students studying the CCNA material

http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080093fd6.shtml

http://www.cisco.com/en/US/tech/tk364/technologies_tech_note09186a0080093f1e.shtml