Jump to content


EIGRP


  • Please log in to reply
71 replies to this topic

#31 martinlo

martinlo

    V.I.P. Member

  • Global Moderators
  • PipPipPipPipPip
  • 2697 posts
  • Gender:Male
  • Location:Land Of 10,000 Lakes

Posted 20 April 2009 - 11:07 AM

what the difference between Passive and Active!!
can anyone explain for me this:

Passive means that the route is in the passive state, which is a good thing because routes in the active state (A) indicate that the router has lost its path to this network and is searching for a replacement.

Edited by martinlo, 20 April 2009 - 11:08 AM.

  • 0

#32 martinlo

martinlo

    V.I.P. Member

  • Global Moderators
  • PipPipPipPipPip
  • 2697 posts
  • Gender:Male
  • Location:Land Of 10,000 Lakes

Posted 21 April 2009 - 12:58 AM

hey again
mean any route RD is less than FD = feasible successor
and any route FD= to the FD mean 10.1.6.3 FD = to 2172416 FD is the successor route?
Right?
one more thing the feasible successor won't be also the backup as the successor route?
Thanks


Earlier I suggested that 1794560 is compared with 2172416.
Correction: 1794560 is compared to 2684416

Example:
P 10.1.3.0/24, 1 successors, FD is 2172416
via 10.1.6.3 (2172416/28160), Serial0/1
via 10.1.4.2 (2684416/1794560), Serial0/0
  • 0

#33 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 21 April 2009 - 07:01 PM

Earlier I suggested that 1794560 is compared with 2172416.
Correction: 1794560 is compared to 2684416

Example:
P 10.1.3.0/24, 1 successors, FD is 2172416
via 10.1.6.3 (2172416/28160), Serial0/1
via 10.1.4.2 (2684416/1794560), Serial0/0


Now you got it
  • 0

#34 martinlo

martinlo

    V.I.P. Member

  • Global Moderators
  • PipPipPipPipPip
  • 2697 posts
  • Gender:Male
  • Location:Land Of 10,000 Lakes

Posted 22 April 2009 - 10:23 PM

My First Post was Correct, not 2nd !
1794560 is compared with 2172416, if it less it is Feasible Successor. Final Answer! Sorry for misunderstanding!

Example: EIGRP
P 10.1.3.0/24, 1 successors, FD is 2172416
via 10.1.6.3 (2172416/28160), Serial0/1
via 10.1.4.2 (2684416/1794560), Serial0/0
  • 0

#35 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 26 April 2009 - 10:17 PM

QUOTE (martinlo @ Apr 22 2009, 03:23 PM) <{POST_SNAPBACK}>
My First Post was Correct, not 2nd !
1794560 is compared with 2172416, if it less it is Feasible Successor. Final Answer! Sorry for misunderstanding!

Example: EIGRP
P 10.1.3.0/24, 1 successors, FD is 2172416
via 10.1.6.3 (2172416/28160), Serial0/1
via 10.1.4.2 (2684416/1794560), Serial0/0


No problem , this is one of the tricks inside the EIGRP , if you understood it then its time for you to ROCK
  • 0

#36 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 30 April 2009 - 08:28 AM


Hello Guys , any questions regarding this Topic : )
  • 0

#37 dgharami

dgharami

    Newbie

  • Members
  • Pip
  • 22 posts
  • Gender:Male
  • Location:India

Posted 29 May 2009 - 08:00 PM

DarkFiber's EIGRP_CCNA in PDF Format

Attached File  EIGRP_CCNA.pdf   1.04MB   149 downloads

smile.gif
  • 0

#38 anderson2005

anderson2005

    Member

  • Members
  • PipPip
  • 131 posts
  • Gender:Male
  • Location:England

Posted 03 July 2009 - 09:41 AM

QUOTE (DarkFiber @ Apr 29 2009, 11:28 PM) <{POST_SNAPBACK}>
Hello Guys , any questions regarding this Topic : )

Yes please...
I am not able to understand the route summarisation with eigrp properly. I am also struggling the ip default-network command with Eigrp. I am not able to understand under which circumstances these two ( "router summarisation" and "ip default-network" ) should be used.

I would really appreciate any help in this matter.



  • 0

#39 surnar

surnar

    Newbie

  • Members
  • Pip
  • 13 posts

Posted 25 July 2009 - 05:42 AM

I have a eigrp static route redistribution between 2 autonomous systems question wondering if i could post it here or guide me into which forum i should.

TIA



  • 0

#40 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 25 July 2009 - 07:18 AM

QUOTE (surnar @ Jul 24 2009, 10:42 PM) <{POST_SNAPBACK}>
I have a eigrp static route redistribution between 2 autonomous systems question wondering if i could post it here or guide me into which forum i should.

TIA



Your more than welcome smile.gif

wating your question
  • 0

#41 surnar

surnar

    Newbie

  • Members
  • Pip
  • 13 posts

Posted 25 July 2009 - 08:19 PM

HELP !!!!

i am trying to propagate the default route on HQ to WEST & W1.

ip route 0.0.0.0 0.0.0.0 Serial1/3


it gets propagated to WEST but is not redistributed to WI
can someone indicate what i am missing out/doing wrong.

i am attaching the image of the topology, the running configs & the sh ip route

TIA






Attached Files


  • 0

#42 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 25 July 2009 - 10:17 PM

QUOTE (surnar @ Jul 25 2009, 01:19 PM) <{POST_SNAPBACK}>
HELP !!!!

i am trying to propagate the default route on HQ to WEST & W1.

ip route 0.0.0.0 0.0.0.0 Serial1/3


it gets propagated to WEST but is not redistributed to WI
can someone indicate what i am missing out/doing wrong.

i am attaching the image of the topology, the running configs & the sh ip route

TIA



router eigrp 100
default-information allowed out
end
wr mem

Edited by DarkFiber, 25 July 2009 - 10:18 PM.

  • 0

#43 surnar

surnar

    Newbie

  • Members
  • Pip
  • 13 posts

Posted 27 July 2009 - 10:02 PM

<!--quoteo(post=851417:date=Jul 25 2009, 01:19 PM:name=surnar)--><div class='quotetop'>QUOTE (surnar @ Jul 25 2009, 01:19 PM) <a href="index.php?act=findpost&pid=851417"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->HELP !!!!

i am trying to propagate the default route on HQ to WEST & W1.

ip route 0.0.0.0 0.0.0.0 Serial1/3


it gets propagated to WEST but is not redistributed to WI
can someone indicate what i am missing out/doing wrong.

i am attaching the image of the topology, the running configs & the sh ip route

TIA<!--QuoteEnd--></div><!--QuoteEEnd-->


router eigrp 100
default-information allowed out
end
wr mem




thanks for your response,

however on WEST even after adding the command, the default route doesn´t show up on W1.

have added on W1 also
router eigrp 100
default-information allowed in

makes no difference.
maybe i have misconfigured something else that is preventing the redistribution of the static route.

TIA
  • 0

#44 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 28 July 2009 - 09:50 AM

thanks for your response,

however on WEST even after adding the command, the default route doesn´t show up on W1.

have added on W1 also
router eigrp 100
default-information allowed in

makes no difference.
maybe i have misconfigured something else that is preventing the redistribution of the static route.

TIA


i used this before & it worked, ok, if your in a hurry try the default-network cmd
  • 0

#45 surnar

surnar

    Newbie

  • Members
  • Pip
  • 13 posts

Posted 28 July 2009 - 08:30 PM

i used this before & it worked, ok, if your in a hurry try the default-network cmd


read all the ref material re EIGRP on the cisco site as well as other ref material on the web.
have tried all the configuration suggestions - no luck

finally on the WEST router added

ip route 0.0.0.0 0.0.0.0 s1/0


and now the static route gets distributed to W1.

thanks for your help.
i would still like to figure out if there is another way to redistribute a "learned default static route" from WEST to W1
  • 0

#46 surnar

surnar

    Newbie

  • Members
  • Pip
  • 13 posts

Posted 28 July 2009 - 08:53 PM

i used this before & it worked, ok, if your in a hurry try the default-network cmd


read all the ref material re EIGRP on the cisco site as well as other ref material on the web.
have tried all the configuration suggestions - no luck

finally on the WEST router added

ip route 0.0.0.0 0.0.0.0 s1/0


and now the static route gets distributed to W1.

thanks for your help.
i would still like to figure out if there is another way to redistribute a "learned default static route" from WEST to W1
  • 0

#47 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 11 October 2009 - 08:13 PM

read all the ref material re EIGRP on the cisco site as well as other ref material on the web.
have tried all the configuration suggestions - no luck

finally on the WEST router added

ip route 0.0.0.0 0.0.0.0 s1/0


and now the static route gets distributed to W1.

thanks for your help.
i would still like to figure out if there is another way to redistribute a "learned default static route" from WEST to W1


Oh yeah, you wont distribute the default unless its in our routing table,
default-info originate always
the always key word make you always send a default route, without the always key word your router wont send teh default route unless it exsist in the RIB.

Thats why it appeared after you configured the default route via s0.

1- default-info originat
2- default network
these are the 2 ways to redistribute default route

there is a 3rd way, but will only send a default route to nei and suppress all other routes
int x/y
ip summary-address eigrp <AS#> 0.0.0.0 0.0.0.0

Edited by DarkFiber, 11 October 2009 - 08:17 PM.

  • 0

#48 yasir vistro

yasir vistro

    Newbie

  • Members
  • Pip
  • 2 posts
  • Gender:Male
  • Location:karachi pakistan

Posted 30 October 2009 - 04:30 AM

hi every body

i need voucher number so kidly help me ...yasir
  • 0

#49 new things

new things

    Advanced Member

  • Members
  • PipPipPip
  • 191 posts
  • Gender:Male

Posted 23 November 2009 - 05:48 PM

i am prepare for exam it is very useful to me
  • 0

#50 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 27 November 2009 - 03:37 AM

i am prepare for exam it is very useful to me


Thanks :)
  • 0

#51 Mr.Mandeep

Mr.Mandeep

    Newbie

  • Members
  • Pip
  • 14 posts

Posted 23 December 2009 - 10:53 PM

The characteristics of EIGRP follow:

"Hybrid routing protocol (distance vector that has link-state protocol characteristics)"


The correct defining property of EIGRP is that It's a advanced distance vector routing protocol, but still distance vector.
All things like, hello packets, event-driven updates. These are all improvements. There are old documents saying, EIGRP is hybrid protocol. In reality, It is a still
DV protocol.

In distance vector, each router receives advertisement from neighbor routers, and applies its local metric, and forward these advertisements to its neighbors containing all routes in routing table. In distance vector, router has no idea which router originated the advertisement.

This is the operation of EIGRP as well, routing being getting advertisements originated by its neighbors. That's what makes EIGRP a still DV protocol.


In link-state, router knows how the complete network looks like - who is connected to whom, and properties of a particular link b/w any two routers.

Edited by Mr.Mandeep, 23 December 2009 - 10:55 PM.

  • 0

#52 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 24 December 2009 - 08:13 PM

The correct defining property of EIGRP is that It's a advanced distance vector routing protocol, but still distance vector.
All things like, hello packets, event-driven updates. These are all improvements. There are old documents saying, EIGRP is hybrid protocol. In reality, It is a still
DV protocol.

In distance vector, each router receives advertisement from neighbor routers, and applies its local metric, and forward these advertisements to its neighbors containing all routes in routing table. In distance vector, router has no idea which router originated the advertisement.

This is the operation of EIGRP as well, routing being getting advertisements originated by its neighbors. That's what makes EIGRP a still DV protocol.


In link-state, router knows how the complete network looks like - who is connected to whom, and properties of a particular link b/w any two routers.


Eigrp is a hybrid as it have some characteristics of DV & Link state.

Examples:
DV = limited hop count ;)
Linkstate= 3 tables routing, topology,nei + keep nei relationship + hellos

Also trust me in this, DV protocols add the metric and then send the update to the nei router, the nei router just put teh update in his routing table.
BUT in Hybrid, the router send its own metric that resides in its routing table , and send this message to the nei, The nei will check the message AND ADD HIS OWN METRIC, and here comes the hybrid protocol with the FD & RD/AD

Thats all :)

SO Eigrp is a hybrid protocol, if that didn't convince you name it an advanced DV.....trust me it wont make a network outage. :P
  • 0

#53 Mr.Mandeep

Mr.Mandeep

    Newbie

  • Members
  • Pip
  • 14 posts

Posted 25 December 2009 - 04:18 AM

Eigrp is a hybrid as it have some characteristics of DV & Link state.

Examples:
DV = limited hop count ;)
Linkstate= 3 tables routing, topology,nei + keep nei relationship + hellos

Also trust me in this, DV protocols add the metric and then send the update to the nei router, the nei router just put teh update in his routing table.
BUT in Hybrid, the router send its own metric that resides in its routing table , and send this message to the nei, The nei will check the message AND ADD HIS OWN METRIC, and here comes the hybrid protocol with the FD & RD/AD

Thats all :)

SO Eigrp is a hybrid protocol, if that didn't convince you name it an advanced DV.....trust me it wont make a network outage. :P



Never mind. Try to read what I said above. You will know.
  • 0

#54 Mr.Mandeep

Mr.Mandeep

    Newbie

  • Members
  • Pip
  • 14 posts

Posted 26 December 2009 - 07:50 AM

Eigrp is a hybrid as it have some characteristics of DV & Link state.

Examples:
DV = limited hop count ;)
Linkstate= 3 tables routing, topology,nei + keep nei relationship + hellos

Also trust me in this, DV protocols add the metric and then send the update to the nei router, the nei router just put teh update in his routing table.
BUT in Hybrid, the router send its own metric that resides in its routing table , and send this message to the nei, The nei will check the message AND ADD HIS OWN METRIC, and here comes the hybrid protocol with the FD & RD/AD

Thats all :)

SO Eigrp is a hybrid protocol, if that didn't convince you name it an advanced DV.....trust me it wont make a network outage. :P



EIGRP has nothing to do with hybrid protocol. The only hybrid protocol commonly used is the multi-area OSPF.

Look, We have to look on a defining property of a protocol, on the fundamental principle on which it operates, and that shall be the decisive thing that tells us what is a LS and what is a DV. Looking on the supporting mechanisms is just looking plain wrong.

We call protocols distance-vector or link-state if they fulfill a certain architectural criteria.

Now, a link-state protocol must work on this basis:


1.) Each router describes in detail its connections to the neighboring objects (routers, networks)
2.) This information is flooded without any change to all routers in an area
3.) Each router consequently has information about all routers, all networks and their detailed interconnections, and from this table, the entire network topology can be drawn
4.) Each router calculates the shortest paths on its own using the topology map

A distance-vector protocol operates by sending a list (an array = a vector) of known networks and their individual metrics. That's about it.

Now, a hybrid protocol is a protocol that represents a network in part by link-state approach and in part by distance-vector approach.

The EIGRP does not do anything similar.

Adding lengths is not a link-state thing, nor is maintaining a feasible distance for a destination. Actually, Dr. Jose Joaquin Garcia-Luna-Aceves, the author of the DUAL, devised the entire DUAL including the feasible distance and the feasibility condition WITHOUT ANY ASSUMPTIONS with regards to DV or LS routing protocols.

The goal of DUAL was actually to avoid having a link-state database to avoid routing loops. If you have a link-state database, the Dijkstra algorithm guarantees by itself that it will not create routing loops. But Dr. Garcia wanted to improve the DV protocols with a supporting conditions how to avoid routing loops without knowing more about the network topology.

Actually, the EIGRP works with the same knowledge about the network as RIP has, it just tries to remember more information about the distance dynamics and it uses some sufficient conditions (the mathematical proofs of which are nontrivial) to be sure that the routing loop will not be created even without knowing how the network looks like.


I hope, It is clear to you now.

Edited by Mr.Mandeep, 26 December 2009 - 08:00 AM.

  • 0

#55 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 27 December 2009 - 09:45 AM

EIGRP has nothing to do with hybrid protocol. The only hybrid protocol commonly used is the multi-area OSPF.

Look, We have to look on a defining property of a protocol, on the fundamental principle on which it operates, and that shall be the decisive thing that tells us what is a LS and what is a DV. Looking on the supporting mechanisms is just looking plain wrong.

We call protocols distance-vector or link-state if they fulfill a certain architectural criteria.

Now, a link-state protocol must work on this basis:


1.) Each router describes in detail its connections to the neighboring objects (routers, networks)
2.) This information is flooded without any change to all routers in an area
3.) Each router consequently has information about all routers, all networks and their detailed interconnections, and from this table, the entire network topology can be drawn
4.) Each router calculates the shortest paths on its own using the topology map

A distance-vector protocol operates by sending a list (an array = a vector) of known networks and their individual metrics. That's about it.

Now, a hybrid protocol is a protocol that represents a network in part by link-state approach and in part by distance-vector approach.

The EIGRP does not do anything similar.

Adding lengths is not a link-state thing, nor is maintaining a feasible distance for a destination. Actually, Dr. Jose Joaquin Garcia-Luna-Aceves, the author of the DUAL, devised the entire DUAL including the feasible distance and the feasibility condition WITHOUT ANY ASSUMPTIONS with regards to DV or LS routing protocols.

The goal of DUAL was actually to avoid having a link-state database to avoid routing loops. If you have a link-state database, the Dijkstra algorithm guarantees by itself that it will not create routing loops. But Dr. Garcia wanted to improve the DV protocols with a supporting conditions how to avoid routing loops without knowing more about the network topology.

Actually, the EIGRP works with the same knowledge about the network as RIP has, it just tries to remember more information about the distance dynamics and it uses some sufficient conditions (the mathematical proofs of which are nontrivial) to be sure that the routing loop will not be created even without knowing how the network looks like.


I hope, It is clear to you now.


Mr Mandeep, i must admit that you got a talent in representing your idea & a very interesting way in documenting it.
We both are talking about the same point, and i agree n all the technical details yur talking abut.

Now, a hybrid protocol is a protocol that represents a network in part by link-state approach and in part by distance-vector approach.
The EIGRP does not do anything similar. (It has, the nei relation ship & 3 tables and hello idea from link-state)

Actually, the EIGRP works with the same knowledge about the network as RIP has. (as you said)


EIGRP, some say its hybrid as of some similarities/features taken from link-state and distant vector, other say its a stand alone protocol......
At the end, this wont make any changes while implementing,operating nor designing it
  • 0

#56 mip104

mip104

    V.I.P. Member

  • Veterans
  • PipPipPipPipPip
  • 4046 posts
  • Gender:Male

Posted 15 April 2010 - 03:19 AM

@dingdong the one who reported the link is not working, I have already downloaded it and it works :) nothing wrong with the link
  • 0

#57 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 17 April 2010 - 07:43 PM

@dingdong the one who reported the link is not working, I have already downloaded it and it works :) nothing wrong with the link


Thank god :)


I'm in the middle of my honey moon, 5 days with no internet...couldn't resist more ;)
  • 0

#58 MaX B

MaX B

    Newbie

  • Members
  • Pip
  • 1 posts

Posted 24 April 2010 - 01:56 AM

Thank you, great work :)
  • 0

#59 JaesonV

JaesonV

    Newbie

  • Members
  • Pip
  • 19 posts

Posted 02 June 2010 - 09:46 PM

EIGRP


EIGRP Summary

 You can factor ... MTU ... into the metric.


I don't know why Cisco got messed up with their docs on this.

We cannot factor MTU into the EIGRP Composite Metric. The K values are just multipliers/modifiers in the Composite metric calculation. Changing the MTU does not impact the distance reported by EIGRP. But I'd agree that MTU is part of the Vector metrics that EIGRP reports for a known network.

I'm not agreeing with that statement because of this other statement -

The MTU doesn’t actually used in the Metric calculations...


Most would mention that the K values represent BW, Load, Delay, Reliability, and MTU. K1 is not Bandwidth as K2 is not Load. It's evident in the formula itself. K5 is definitely not MTU.

But still a nice summary. :P

Edited by JaesonV, 02 June 2010 - 09:47 PM.

  • 0

#60 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 07 June 2010 - 05:14 AM

I don't know why Cisco got messed up with their docs on this.

We cannot factor MTU into the EIGRP Composite Metric. The K values are just multipliers/modifiers in the Composite metric calculation. Changing the MTU does not impact the distance reported by EIGRP. But I'd agree that MTU is part of the Vector metrics that EIGRP reports for a known network.

I'm not agreeing with that statement because of this other statement -

Most would mention that the K values represent BW, Load, Delay, Reliability, and MTU. K1 is not Bandwidth as K2 is not Load. It's evident in the formula itself. K5 is definitely not MTU.

But still a nice summary. :P


Yes the MTU is not particpating in the Metric calculation, but its send in the update.
You cant play with the MTU in the metric calculations as per teh below.

The composite metric for each EIGRP route is calculated as

EIGRP metric = IGRP metric * 256
IGRP metric = [k1*BWIGRP(min) + (k2* BWIGRP(min))/(256-LOAD) + k3*DLYIGRP(sum)]
x [k5/(RELIABILITY + k4)]
If k5 is set to zero, the [k5/(RELIABILITY+k4)] term is not used.

Given the default values for k1 through k5, the composite metric calculation used by IGRP reduces to the default metric:
IGRP metric = BWIGRP(min) + DLYIGRP(sum)
BWIGRP(min) = 107/BW(min)
DLYIGRP(sum) = DLY(sum) /10
  • 0





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users