Networks Routing question

steviejay

Now in Black and White!!
Joined
Jun 9, 2002
Messages
3,343
Location
Glasgow, Scotland
ok I know this is going to be difficult for people if you don't know what I'm talking about but I've got an exam tomorrow and I'm struggling, mainly due to my comtempt for my the complete muppet of a teacher, anyway. I'm doing past papers in an attempt to prepare mtself and I have a question which I have no idea how to do this question and was wondering how it's done-

A router J on a WAN receives delay vectors from it's neighbours as follows:

E 10 7 12 18 0 11 6 6 8 6 9 7
F 8 14 9 5 9 0 16 8 6 12 16 14
H 9 5 14 8 12 10 9 0 7 12 6 12
L 4 7 9 11 5 10 8 6 12 4 5 7

J has estimated it's delay to it's neighbours E, F, H and L as 8, 6, 7 and 12 msec respectively. Calculate the new routing table for Router J.

Can anyone make sense of this?
 
A complete guess ...

I guess the 12 numbers next to each neighbour node are the delays those nodes measure to 12 destination nodes labelled A, B, C, ... L. For E, F and H the delays to themselves are correctly given as zero, but I'm not sure why L would report 7 mSec to itself.

If this assertion is correct, then presumably you can find node J's total delay to each of the 12 destinations via each of its neighbours by adding the delays. For example:

Delay from J to A via E = 18 mSecs, as it's J->E (8 mSecs) + E->A (10 mSecs)

Similarly:
J to A via F is 8 + 6 = 14 mSecs
J to A via H is 7 + 9 = 16 mSecs
J to A via L is 12 + 4 = 16 mSecs

So the fastest route to A is via F (14 mSecs), and J should define F as its next hop to route traffic to A.

Repeat for all 12 destinations to get a full routing table. A spreadsheet would do it faster than the manual method. Put the delay vectors into a 12 column x 4 row table and add the first hop delays for E, F, H, L to their respective rows, then look for the minimum total delay in each column.

I hope this doesn't just confuse the issue further for you :)
 
the answer is this (bear with me, numbers a plenty)

10 8 9 4 + (8 6 7 12) gives 20 14 16 16 a f 14msecs (you were right woo)
3 14 5 7 + (8 6 7 12) gives 15 20 12 19 b h 12
12 9 14 9 + (8 6 7 12) gives 20 15 21 21 c f 15
18 5 8 11 + (8 6 7 12) gives 26 11 15 20 d f 11
0 9 12 5 + (8 6 7 12) gives 8 15 19 17 e e 8
11 0 10 10 + (8 6 7 12) gives 19 6 17 22 f f 6
6 16 9 8 + (8 6 7 12) gives 14 22 16 20 g e 14
6 8 0 6 + (8 6 7 12) gives 14 14 7 18 h h 7
8 6 7 12 + (8 6 7 12) gives 16 12 14 24 i f 12
6 12 12 4 + (8 6 7 12) gives 14 18 19 16 j e 14
9 16 6 5 + (8 6 7 12) gives 17 22 13 17 k h 13
7 14 12 0 + (8 6 7 12) gives 15 20 19 12 l l 12

Now, the a f 14 must mean the lowest time between routers a and f is 14msecs as it's the lowest of the 4 numbers.

the (8 6 7 12) is the delay from the neighbours of J which is given in the question, I'm just not sure how the other numbers are attained. if that makes sense. I mean how can 10 8 9 4 + (8 6 7 12) give you 20 14 16 16?
 
that's just confusing, you're logic makes sense (thank you very much by the way) and the majority of your calcs make sense and even agree with what I've got. but the 20 confuses the hell out of me. Unless of course it is actually 18 and the lecturer just balls it up, which normally I would dismiss be it any other lecturer but this guy is a walking talking example of why there should be warning stickers on things, hmm, a potential mistake.... what a tool
 
my god....... he's buggered up the results...... that's why it never made any sense to me....... dude, you were right, you commented upon the fact that L reports a delay of 7 mSec to itself.... it's that because he done it wrong. I just noticed it all. mother of god I've been so stupid. He's messed up the whole thing....

I'd call him a tool but I can't do that, tools have a purpose...... he's just a waste of space (he's alot of things, but this being a family forum and all...)
 
Back
Top Bottom