Network+ subnetting query

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 all,

I'm looking through Osbournes CompTIA Network+ Certification Study Guide, Fourth Edition (Glen E. Clarke), "Chap 5 - Subnettng and Routing", and have come across something that just doesn't site right with me.

He shows that given a 10.0.0.0 network (with the default class A subnet mask of 255.0.0.0), to divide this into two possible subnets you would use the mask 255.192.0.0, stating that the possible networks 10.0.0.0 and 10.192.0.0 are invalid as the result is either all 1's or all 0's in the second octet for the net portion of the address.

However, RFC 1878 (written in 1995! - and they way that I have always understood it) abolished this practice meaning that to split the 10.0.0.0 network into two usable subnets, we can use the mask of 255.128.0.0, resulting in two network IDs of 10.0.0.0 and 10.128.0.0

The mask as given in the book would therefore result in 4 networks, not 2!!!

But what does CompTIA say about this?

 

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.

You are correct that most network devices these days can use any subnet. However, some older routers still use those "old rules" regarding the all-zeros and all-ones subnets, so it's good to know the theory behind it.

As for what CompTIA expects, who knows? That said, I have not yet seen a question from CompTIA, Microsoft, or Cisco that would leave you to doubt what the correct answer is on an exam... meaning, they won't typically try to trick you by giving you both the n and n-2 subnet calculations as possible correct answers.

Not saying that they absolutely won't put questions like that on the exam, because CompTIA recruits just about anyone who's passed an exam to write questions for them... just saying that they I haven't seen any misleading subnet questions from them in the past.

 

I believe most networks can handle subnet IDs being all ones or all zeros, but the internet powers-that-be say they shouldn't use them. For the purposes of the N+ the correct subnet mask is 255.192.0.0.

 

Bloody ridiculous! I know that CompTIA are a little behind the times but I though that this update to the exam to at least take on new (if you can call 15 years old new!) recommendations and practices.

Now wonder people get real confused about all this stuff. At working in Networked environment for the last 10 years I have yet to see such subnetting where the lowest and highest possible network are ignored.

 

But the problem is this: the networks that you'll be supporting might not support those subnets. You might not have seen them; I have.

So let's say you studied for the "new way" only, where any subnet can be used. Then you try to implement that on a production network... and it doesn't work. You'd be left wondering why that happened, wouldn't you? After all, you wouldn't have been trained on the "all-zeros" and "all-ones" subnet restrictions, so you wouldn't have any clue about that limitation. Whereas if you learn things the "old way", then the scenario I mentioned won't surprsie you; you'll know the reason why.

Sometimes knowing the "old school" way IS best for helping you troubleshoot problems, even if you don't use the technology. I'm surprised that you don't already know that by now after working on networks for 10 years. ;)

 

As you say, maybe its the area I'm working in. This has been mainly small networks and now university/large scale education networks. In both of these scenarios normally new equipment is employed and the best foot forward utilises newer, not older, technologies.

It just a tad confusing when you finally have to hit some books after just doing what you do for such a long period of time, only to find that old ways are still de rigour.

 

Now check this vid out, section 1.4 Part 2, which show the subnetting method I am accustomed to :-

www.freenetworkplus.com

I understand binary maths and networking quite well, but these different procedures when relating to an exam are very poor. Now wonder lots of people can't get there head around this issue.

 

True... but what do you expect for free?

 

The way I see it, the n-2 rule works on any piece of kit, the n rule should work on any piece of kit but might not. Therefore no harm will come from teaching inexperienced peeps like myself to use the sure-fire way of doing it and avoiding confusing the issue. This is after all an entry-level cert so I doubt anyone's going to be beating down my door cos they need a network designed in the near future.

 

Exactly my point. Repped!

 

Well, would believe it. I've just sat the exam and got one of these questions, but I'm pretty sure I answered it incorrectly.

I answered it the way I would normally do things without even thinking about it.

....
....
....


Just goes to show that its very difficult to undo what you know, especially if its the only way you have done things, and have done things that way for a long time.

Still, 840/900 ain't bad, especially considering I didn't start going through the books till Monday just gone.


EDIT - just though my answer might be right as the low subnet would still not be legal - Oh I don't know...

 

Just read the same chapter in Meyers book 2009 book, which is somewhat different to what many of the Microsoft books tell you these days. Oh well, better safe than sorry and exclude any 0's and 1's from being subnets I guess...

 

Your answer would have been right; as you have figured out, 10.10.0.0/21 wouldn't have worked under the old rules either.

Like I said, they're not going to give you the "old rule" correct answer AND the "new rule" correct answer and ask you to choose between them; they're going to give you one OR the other. In this case, they gave you something that couldn't possibly EVER have worked under the "old rules"... leaving only one possible set of rules to go by.

You might want to remove that test question, just in case the "CompTIA police" are scanning the forums for NDA violations. Better safe than sorry.

Congrats!

 

I did make changes to both question and numbers so that it bore no resemblance to the actual question, but I do get your point

Maybe I should have started this Certification thing years ago. Even though I should have been able to complete this exam with no problems, I have still learnt some new things - even if it may not be applicable to my particular field.

 

I agree. I learned things when writing an A+ practice exam after being in IT for 5 years (and "messing with computers" for 23 years). Certifications are worth doing.