Sebnet zero and the all 1's subnet.

Oh man! Ad-blocking software has been detected! :'(

This website is run by the community, for the community... and it needs advertisements in order to keep running. Blocking our ads means your killing our stats!
Please disable your ad-block, or become a premium member to hide all advertisements and this notice.

Hi guys

I've got my CCNA Intro exam Monday. I need to get the whole CCNA done by November 20th for me to qualify for a promotion Ive been offered. I've worked with Cisco stuff for years so it's about time I got qualified

Anyway, my question is on subnet zero and the all 1's subnet. I'm confident with my subnetting (as long as I remember to RTFQ) but I am unsure about how to treat subnet zero and the broadcast subnet in terms of the CCNA exam.

From what I can gather, in the CCNA exam you should treat all questions in the terms of using these subnets unless the question specifies that it is not safe to use them.

For example if the question is how many subnets and how many hosts would the subnet address 150.150.0.0 with a /18 subnet mask provide, then the answer would be:

1) 4 subnets and 16,382 hosts.

However if you are not using subnet 0 and the all 1's subnet the answer would be:

2) 2 subnets and 16,382 hosts.

Am I safe in thinking that unless the question in the CCNA Intro exam specifically says that you are not to use subnet 0 (and therefore not the all 1's subnet?) then I am safe with answer 1?

And also am I correct in thinking that if you are told not to use subnet 0, then this also means not using the broadcast subnet also?

Thanks

Daz

 

Oh man! Ad-blocking software has been detected! :'(

This website is run by the community, for the community... and it needs advertisements in order to keep running. Blocking our ads means your killing our stats!
Please disable your ad-block, or become a premium member to hide all advertisements and this notice.

EvilDaz,

Assuming you pass INTRO next monday, in order to earn the CCNA:

1. you will have to take and pass ICND up to and including 6 November 2007.
2. Take and pass the ICND2 after 6 November 2007.

That is because Cisco has set a last day to test of 6 November 2007 for the following exams:

1. 640-801 CCNA
2. 640-822 INTRO
3. 640-811 ICND

Source:

1. CCNA - Career Certifications & Paths - Cisco Systems - http://www.cisco.com/web/learning/le3/le2/le0/le9/learning_certification_type_home.html

 

EvilDaz,

What class is the IP address 150.150.0.0?

 

For CCNA assume that you are NOT to use subnet zero nor all 1's subnet unless the question specifically says to do so, or infers so: this is the advice in the Cisco Press Exam Guides - whether this will apply to the new exams remains to be seen.

 

Hi

Thanks for all your input.


>>r.h.lee - Yeah, I new the deadline for these exams were in November, for some reason I had it in my head it was the end of November, oh well looks like I better get a move on .The class of the address is a class B.


>> MacAllan - Thanks, but I'm still not 100%. Other places say that as ip subnet-zero is enabled by default on all modern cisco equipment then you should treat all questions as if allowed to use subnet zero unless the question specifically says not to use subnet zero, not vice versa like you suggest.

Is what you say for certain as I really need a definitive answer before taking any exam.

Also I am still unclear that if you are not to use subnet zero does this implicitly impy that the all 1's subnet is also unuseable too?

Thanks

Daz

 

To be honest, I have yet to see a Cisco exam question where it is unclear whether they want you to use the all 0's or all 1's subnet. They'll either not have that option, they'll have that option so it's the only one that makes sense, or they'll come right out and say subnet zero is on/off. In any case, it should be clear. At least, that's been my experience.

 

According to the forums at CISCO (sign up and do a search for "subnet 0") a post was made saying that due to the confusion regarding whether it's ok to count subnet zero or not they will indicate their preference in the exam question.

 

Thanks BosonMichael and Liqua. That has put my mind at ease, I can get on with it now.

All your input is much appreciated guys.

Thanks

Daz

 

EvilDaz,

Ok, which routing protocol is being used?

 

r.h.lee

I'm not sure I get you?

The question was purely an example of working out the number of subnets and hosts you get from a class B network when using a particular subnet mask. The network address was completely random and made up off the top of my head and the mask I used was one that would give you 4 (2 if not using the all 0's and all 1's) subnets.

This itself is actually irrelevant in such that it is only serving as an example to whether I should answer any question in the CCNA exams, where the amount of subnets is asked, in terms of using subnet 0 and the all 1's subnet if the question does not specifically say whether subnet 0 is allowed or not.

There is no question of routing or routing protocols.

But thanks for your time anyway

Daz

 

I don't know how specific you want it: it's the advice from the Cisco Press Exam Guide

I took that advice for the exam I sat this year, and didn't regret it. Not all questions made it clear what was wanted, and in those cases I assumed I was intended to NOT use subnet zero. I know where I was marked down, and it wasn't those questions.

Yes, most real world cases use the router command to enable subnet zero, but the official cisco recommendation even now, is that although it is possible, it isn't recommended.

For CCNA, assume no, for CCNP, I think the advice is to assume yes.

Also, if you don't use subnet zero, you certainly don't use all 1's.

But like I said, how this will work for the new exams, I have no idea.

 

EvilDaz,

Actually, I believe that it is highly relevant since the way you subnet is dictated by the routing protocol. In the case of a layer 2 switched LAN without routing then you use classful subnetting. In the case of the distance vector routing protocols RIPv1 and IGRP, you also use classful subnetting. In the case of link state routing protocol, you're going to need to use classless subnetting. In the case of a hybrid routing protocol, you may use either classful or classless subnetting. That's why I'm asking which routing protocol.

 

Actually, subnet zero is now enabled by default on Cisco routers, as of IOS 12.0. Use of the all-zeros and all-ones subnets was discouraged in the past, but that's no longer Cisco's stance. Even RFC 1878 states that the practice of using these subnets is obsolete.

From Cisco:

 

Routing protocol is irrelevant as to whether one uses the all-zeros or all-ones subnets. You can subnet using RIPv1 and IGRP, but the entire routed network must use the same mask: link

 

What exactly do you mean by classful subnetting? How would I subnet a Class C network address classfully??

 

MacAllan,

To me, classful subnetting is when you take the default mask for Class A, Class B, or Class C depending on the class of the IP network address given, then you borrow a specific number of bits from the host portion to create subnets. This is in contrast with Variable Length Subnet Masking (VLSM).

 

It makes things difficult to follow if you are using definitions for things that are non-standard. VLSM means that the subnet mask is different in different parts of the network - for example a WAN connection has a /30 mask, where a LAN subnet of the same network address might have a /26 mask. It has nothing to do with classful or classless networking. Subnetting a classful address introduces classless, there is no such thing as a classful subnet.

As BM has already posted, RIPv1 and IGRP deal perfectly well with classless addressing: but they can not cope with VLSM.

 

MacAllan,

So what do you call subnetting using the same subnet mask in the whole network such as a RIPv1 network?

Classful or classless IP subnetting is related to classful or classless routing protocols. You can't use classless IP subnetting in a classful routing protocol such as RIPv1. You also can't use classful IP subnetting in an OSPF network because OSPF can't automatically summarize.

I disgree that "Subnetting a classful address introduces classless,..." Classful IP subnetting is based on the default mask for the Class of IP address then borrows bits from the host part of the IP address to create subnets. Classless subnetting is based on the assigned network or subnetwork then borrowing bits from the host part of the IP address to create variable subnets for various purposes such as LANs, VLANs, and WAN links. Under classful IP subnetting, you have to use the same subnet mask for LANs, VLANs, and WANs.

Please do not misquote someone else. BosonMichael clearly stated "You can subnet using RIPv1 and IGRP, but the entire routed network must use the same mask: link." He said "subnet" not "classless addressing." Here's an update: VLSM = classless IP addressing. So what do you mean by "...deal perfectly well with classless addressing: but they can not cope with VLSM."?

 

I think both of you are trying to say similar things, and I think you both know "how it works", but I think there's just a difference in terminology. But search cisco.com: you won't find a single reference to "classful subnetting" or "classless subnetting". Cisco doesn't call VLSM "classless subnetting", nor would I.

VLSM can be used only with classless routing protocols, but subnetting itself is neither classful or classless. Subnetting is subnetting. How routing protocols handle addresses is classful or classless. RIP is classful, which means it advertises networks at the major bit boundary. But that has nothing to do with whether you can subnet or not, and we shouldn't confuse the issue by calling it "classful subnetting". Classful means only that you cannot use variable-length subnets throughout the network (VLSM) because of how RIPv1 handles advertisements.

On the other hand, RIPv2 is classless... meaning, it doesn't assume that networks should be summarized at the major bit boundary; it advertises subnet mask information in the routing updates. And thus, you can variably subnet your network.

You can certainly subnet OSPF using normal classful masks, like /8, /16, or /24. You can subnet your OSPF network however you like... on the classful bit boundary or outside of it. Thus, the term "classful subnetting" is confusing, and that's why it's not used by Cisco.