12-29-2009 09:25 PM - edited 03-04-2019 07:05 AM
Hi Experts,
In Eigrp I know that the following contributes the Metric
1. Bandwidth
2. Delay
3. Load
4. Reliability
5. MTU
I trying to understand how MTU influence the Metric. It labbed it. R1 (ser
0/0) ----(ser 0/0) R2
I changed the metric of R1 ser 0/0 to 1200 and noticed the Metric value
didn't change. How and when MTU comes into play.
Also I noticed the Eigrp messages donot contain MTU details.
Can you pl give your valuable time and help us
sairam
Solved! Go to Solution.
12-30-2009 12:02 AM
Hello Sairam,
this is a good question.
MTU is not part of the composite metric. However, EIGRP routes both internal and external are carried within TLV data structures.
The TLV is oriented at providing in multiple fields data ready for EIGRP calculation of local node that is:
prefix
prefix len
minimum BW on path
cumulated delay on path
and also:
minimum MTU on path
number of router hops
you can see this information using sh ip eigrp topology
the local node to calculate the metric from its own point of view just needs to:
add the delay of its own interface where the update has been received
to verify the bandwidth parameter of its own interface where the update has been received if its higher of current minimum BW this field is left unchanged.
to add 1 to router hop field
to verify the MTU parameter of its own interface where the update has been received if its higher of current minimum MTU this field is left unchanged.
MTU and number of router hops are used as tie breakers in case of paths with equal EIGRP metric, the path with the highest minimum MTU is preferred as well as the path with the minimum router hop count.
to be noted that delay already provides a measure of hop count weighted with delay of each link in the path.
Hope to help
Giuseppe
12-30-2009 12:02 AM
Hello Sairam,
this is a good question.
MTU is not part of the composite metric. However, EIGRP routes both internal and external are carried within TLV data structures.
The TLV is oriented at providing in multiple fields data ready for EIGRP calculation of local node that is:
prefix
prefix len
minimum BW on path
cumulated delay on path
and also:
minimum MTU on path
number of router hops
you can see this information using sh ip eigrp topology
the local node to calculate the metric from its own point of view just needs to:
add the delay of its own interface where the update has been received
to verify the bandwidth parameter of its own interface where the update has been received if its higher of current minimum BW this field is left unchanged.
to add 1 to router hop field
to verify the MTU parameter of its own interface where the update has been received if its higher of current minimum MTU this field is left unchanged.
MTU and number of router hops are used as tie breakers in case of paths with equal EIGRP metric, the path with the highest minimum MTU is preferred as well as the path with the minimum router hop count.
to be noted that delay already provides a measure of hop count weighted with delay of each link in the path.
Hope to help
Giuseppe
12-30-2009 01:38 AM
Giuseppe
As usual an excellent detailed reply. Logged on this morning and see your 4 short and i couldn't resist
Congratulations, the forums are a much better place for you being involved in them.
Jon
12-30-2009 03:52 AM
Hello Jon,
I can only say thanks for your very kind words and for your help in reaching this new status!
Sometimes I wonder if we are too much involved in these forums, and I know that the answer is yes.
Some of my job colleagues say that I have a second job here...
Best Regards
Giuseppe
12-30-2009 04:09 AM
Sometimes I wonder if we are too much involved in these forums, and I know that the answer is yes.
I think you may be right but then it beats watching tv or playing games in my opinion. I tend to think of it as one of my main hobbies.
With the added benefits of keeping your skills up to date and just as importantly helping people out when they are stuck for a solution. I'm sure we have all been in the position where we are banging our heads against the wall trying to understand how to do something and to be able to help people out makes it all worthwhile.
Plus it's nice to get a star or 2
Jon
12-30-2009 05:03 AM
Hi all,
Giuseppe, you have my most sincere congratulations! I could see this coming yesterday. I wanted to help like Jon did, but in some of your latest posts I didn't have the necessary knowledge to judge your posts. In reading one of those posts I ended up studying 7600 datasheets at 4 a.m. in the morning!!! Before deciding to finally go to sleep I thought "hey, he's gonna make it anyway, doesn't need my help"! And also thought "I'll just write some code. When point threshold is crossed, an alarm will go off and I will know it's time to see the star changing in front of my eyes". I envy Jon for stealing my thunder!
You guys are more involved than I am. Still, I totally agree with Jon's view about this hobby and the rest. I am more impressed with how technology works rather than the end user services of the internet. If I have some free time, I really don't know what to do with the internet! Besides, in these forums there are many more opportunities for fun than it seems at a first glance!
Keep up the good work and have fun!
Best Regards,
Maria
12-30-2009 05:11 AM
Maria / Giuseppe
Before deciding to finally go to sleep I thought "hey, he's gonna make it anyway, doesn't need my help"!
Couldn't agree more. Giuseppe didn't need my help whatsoever, i just couldn't resist
I envy Jon for stealing my thunder!
Jon
12-30-2009 03:57 AM
Hi Giuseppe
Hope you are doing well. Merry Chirsmas and adv. New year wishes
After a short gap I am seeing you again.Thanks for sharing your valuable ideas
Infact I tested this in lab setup and using wireshark I captured the hello packet. But I couldnt trace any MTU parameter there. That is the seed of this post.
I googled enough to get the clue. But no use
However since you are said the reason for MTU involved in path decision, I will record it and go ahead.
I have rated it as extremely helpful and question answered
see you again
sairam
12-30-2009 04:00 AM
Hello Sairam,
just a little add:
the router hop count has also another use: in EIGRP there is (or there was?) a max hop count concept that has default values of 100 (or 224?) that I never see in action but were in some test questions when I was studying for CCNP.
These kind of protocol details are reported in books like:
Routing TCP/IP volume I by Jeff Doyle and Jennifer Dehaven Carroll that is a must
you need to capture the update packets to see the TLV of prefixes they are not sent in hellos
Edit:
I wish you an Happy New Year and peaceful
Hope to help
Giuseppe
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: