Dirty subnetting again

borsky

Here is another example from Transcender that I have doubts about:

There are 3 subnets and 2 computers on each subnet on a routed network (comp1 and comp2 on subnet 1, comp2 and comp3 on subnet2 etc)
Computers on all subnets are configured with 19-bit subnet masks(I only indicate the last three bits in the subnet ID):

Subnet 1: 101
subnet 2: 100
subnet 3: 110

Comp1 on subnet 1 by mistake is configured with 18 bit subnet mask.
Question: wich computers can comp1 communicate with?

Transcender suggests that comp1 identifies subnet 1 and subnet 2 as local because their first 18-bit is the same, and subnet 3 as remote. Therefore comp1 can communicate with computers on subnet1 and and subnet3, but not subnet 2.

Now I think it is wrong because comp1 has 18-bit subnet mask and it cannot comunnicate with comp2 which has 19-bit subnet mask simply because they are not configured to be on the same subnet. Hosts cannot identify other hosts as local if their subnet masks are different.
Comp1 therefore cannot communicate with any subnets because its default gateway with its 19-bit subnet mask is on a different subnet too.

Source: question B10

Webmaster

Seems like a tricky one, but is rather simple.... and definitely correct:

Computer1 doesn't care about the subnet mask of other hosts. It uses its own subnet mask to determine which IP addresses (not ipaddress/subnetmask combi's)are on the same local subnet. (the subnetwork range...)

The IP of computer 1 is in the subnet 147.60.128.0 - 147.60.191.255, computer 1 considers computer 2, 3 and 4 and router1 to be in the same subnet as well, which is only true for computer 2 and the router....

Computer 1 cannot communicate with the computers in subnet 2 because he thinks they are on the local network, but can communicate with every other.

(side note: if you have two computers you can simulate the situation of comp1 and comp2 having the different subnet mask, but being in the same 'network range' and still able to communicate with eachother.)

Johan

borsky

Thanx, I tried and it works.

But as far as Transcender is concerned I just don't trust it anymore.
In their explanation they contradict themselves and as well as you:

"...Local addresses are those that have the same network ID as the local computer. A network ID is defined by using a subnet mask which specifies how many leftmost bits in an IP address constitute the network ID...."

According to this computer 1 cannot communicate any of the computers.

By the way as a result of my letter to Transcender about that particular question that I quoted in our previous argument on calculating subnet ID, will be updated soon.

Webmaster

borsky wrote:

In their explanation they contradict themselves and as well as you:

I'm not sure where you think I contradict myself (thanks which I don't, neither does Transcender)but reading your previous post, shows YOU are missing the point/logic and accuse me of contradicting myself..

borsky wrote:

...Local addresses are those that have the same network ID as the local computer. A network ID is defined by using a subnet mask which specifies how many leftmost bits in an IP address constitute the network ID...."

According to this computer 1 cannot communicate any of the computers.

According to 'this' computer 1 CAN communicate with the other computers. The subnet mask is of local significance, it's used to determine which IP addresses are on the same subnetwork... computer 1 doesn't care at all about the subnet masks of other hosts... it doesn't even know it... again it cares about which IPs are on the same subnet...
Referring to the Transcender question, just do some calculating and you'll see the question is 100% correct, it is simply a very good question... one of those that shows if you are ready or not and if you understand the concept of subnetmasks and subnetting.... or not

You're welcome,

Johan

borsky

Your point has been proofed I tried.

This time I am not arguing with you, I am arguing with transcender,s explanation which is misleading to those who are not as spot on at subnetting as you are.
"....a network ID is defined by using a subnet mask which specifies how many leftmost bits in an IP address constitute the network ID...."

It means that subnet mask is required to define network ID regardless of local or remote.
I did not know that local host don't care about subnet mask to identify local host and trancender definetly did not help. So thank you for your reply.
But as for the contradiction, well read what you wrote:

"...The subnet mask is of local significance, it's used to determine which IP addresses are on the same subnetwork... computer 1 doesn't care at all about the subnet masks of other hosts..."

subnet mask = determining which IP addresses are on the same subnet
AND
subnet mask = computer1 doesn't know about it and doesn't care about it to determine which hosts are on the same network.

Is it not a contradiction?

Webmaster

borsky wrote:

Your point has been proofed I tried.

But as for the contradiction, well read what you wrote:

"...The subnet mask is of local significance, it's used to determine which IP addresses are on the same subnetwork... computer 1 doesn't care at all about the subnet masks of other hosts..."

subnet mask = determining which IP addresses are on the same subnet
AND
subnet mask = computer1 doesn't know about it and doesn't care about it to determine which hosts are on the same network.

Is it not a contradiction?

You are obviously twisting my words... I'm not saying:"computer1 doesn't know about it and doesn't care about it to determine which hosts are on the same network"

I'm saying: computer 1 doesn't care at all about the subnet masks of other hosts...

It cares al lot about its own subnet mask, again: it uses its own subnet mask to determine in which IP network range it is, when this network range is known it also knows which other ip addresses/hosts fall in that range... (without having to know the subnet masks of those other hosts)

To put it very simple: I am computer1, I have address 7, I know -because of my own subnetmask in which address range I'm in, let's say range 1-10, that means I can contact every computer in this (1-10) range even if one or more of these computers have another subnet mask which make them (the other computers) think they fall in an other range, let's say 6-16 (I could still contact 6,8,9 and 10).

(The above assumes they are on the same physical segment, which is not the case in the Transcender question with some computers...)

I really hope this helps...

If you read carefully and once you'll understand, you'll also see that Transcenders explanation is exactly to the point.
As for "not being as spot on subnetting" This is the kind of level of subnetting knowledge that people should be able to expect from an MCSA... IMHO. The 70-218 exam is not an entry-level exam. Although you might not even get subnetting questions in the 70-218 exam, this is more important for the 70-216 and Cisco's CCNA and CCDA exam.

Don't get me wrong, but have you ever looked at the Network+ certification?

borsky

Dear Webmaster,

you are really helpful, and I found our previous argument helpful too where you firmly stated at the begining that I was wrong but finally it turned out that I was not.

I have done 216 and used transcender which helped me a lot because it was clear. But 218 is not clear at all and I get confused.

You are right about this issue as I stated earlier. What I am trying to say is that the explanation is not clear and can be misunderstood. It is not obvious that the local host is using its own subnet mask to determine who is on the same subnet and don't care about other host's subnet mask at all for doing this.

Finally I did not twist your words the semantic was exactly the same

Take care

Webmaster

borsky wrote:

you are really helpful, and I found our previous argument helpful too where you firmly stated at the begining that I was wrong but finally it turned out that I was not.

I didn't state you were wrong, and now you are implying I was wrong... I explained they used traditional subnetting...
(for those interested www.techexams.net/forums/viewtopic.php?t=447)

Finally I did not twist your words the semantic was exactly the same

I don't agree (neither would a computer )
"computer 1 doesn't care at all about the subnet masks of other hosts..." is something different than "computer1 doesn't know about it and doesn't care about it to determine which hosts are on the same network."

It does care about it (the subnet mask...) just not about the mask of others. All I'm saying is that without your twist there is no contradiction in my words... which I tried to explain in my previous post.... and I won't start again...

Glad to be of help, let me know if you have anything else

Johan

RussS

Hey Borsky

I think there are a couple things you should take into account when discussing subnetting (or any network environment for that matter) when Microsoft and the MCP certifications are being looked at.
First off, the MS network is a fantasy and the real world doesnt work like that at all. We have cisco routers, Unix/Linux servers and a whole lot more that they like to completely disregard
Secondly, when viewing a 70 series question you must always be aware that they are written for a microsoft only network and the correct answer is usually the one that costs you the most in licence fees
From reading your previous posts I seem to get the impression that you will search anywhere to find a site that backs up your idea on how things work. Personally having already passed 215 I am still confused by a lot of subnetting secenarios/questions, but I do know that out in the real world most of those scenarios are completely bogus and if you keep your study material you will have that to draw on if you ever run across a situation you do not understand - or you can post here and Pavlov or Webmaster will come up with a reasonable response

borsky

RussS,

subnetting is more complex then I thought as it turns out from our argument with webmaster. I am a psychologist and new to IT, and only know about Microsoft, and I am only preparing for Microsfoft exams, so apparently very unexperienced in "real world" networking and subnetting.
But I am glad that we can discuss things in here, after all that what this forum is for, and it helps me and I hope it will in the future.