Cisco Tips & Tricks

July 6, 2006

CCIE notes for EIGRP

Filed under: IP Routing, Router, [EIGRP] — ciscotips @ 6:20 pm

At last i have made up my mind to sit for ccie R&S written by early August. I am starting a tips & tricks section for dynamic routing protocols. I will start with EIGRP and then gradually move towards OSPF & BGP.

EIGRP uses DUAL ( Diffused update algorithm). Now what is DUAL? This fancy name is being used by cisco as a kind of marketing tool. Cisco calls EIGRP as a hybrid protocol. It has features of both distance vector & link state protocol. EIGRP topology table will always keep a feasible successor. Feasible successor is nothing but a backup route.

Lets say we have two links. One is 56k & other is a T1 link. Now EIGRP topology table will contain t1 as a primary route and 56k link as a feasible route. The difference between EIGRP and other dynamic routing protocols is that due to the feaible successor already in the topology table,Convergence time is very fast.

Incase of T1 link failiure, in fraction of seconds 56k link will become a primary route for EIGRP. This is the magic of DUAL, in other protocols like OSPF it will take atleast 50 secs to converge as it has to do everything from the scratch.

-EIGRP route metrics :- All of us have known about k values in EIGRP. but what are ‘k’ values?K values are the meterics being used in EIGRP to influence routes. routes in EIGRP can be influenced with 5 different ‘k’ values. k1,k2,k3,k4 & k5. Bandwidth and delay are the default metrics in EIGRP but we have load, reliability and MTU which can also influence routes in EIGRP. Cisco doesn’t recommend enabling other ‘k’ values like load. the reason behind that is that it can increase cpu cycles and thus can slow down the whole process.

IF we have two links with equal cost (bandwidth), both eigrp & ospf will automatically load balance between the two links. Incase we need to prefer one link over other, we will have to change the bandwidth statement ( i.e. we ‘ll lie to the router) to a lower bandwidth so that the other link will be preffered over this link. This practice will definitely resolve our routing problem but it can lead to other problems. bandwidth statement is not only used by routing protocols but also used by qos for evaluating/priortizing traffic. changing bandwidth can have adverse effects.

In eigrp we have a feature called delay we can simply go under interface configuration mode and change delay. its a cool feature as even if bandwidth is same for 2 interfaces, delay will help us in making one link as primary and other secondary by simply increasing value of delay.
Now its a time for EIGRP tip. Not many people know about this command but using this command we can really do wonders in EIGRP.

metric weights tos k1 k2 k3 k4 k5

where k1 is bandwidth, k2 is load, k3 is delay, k4 is reliability and k5 is MTU. This command is used under router configuration mode.
default usage of this commmand is meteric weights 0 1 0 1 0 0. by default only bandwidth and delay is accounted for metric calculation.

Use this command to alter the default behavior of EIGRP routing and metric computation and allow the tuning of the EIGRP metric calculation for a particular type of service (ToS).

If k5 equals 0, the composite EIGRP metric is computed according to the following formula:

metric = [k1 * bandwidth + (k2 * bandwidth)/(256 – load) + k3 * delay]

If k5 does not equal zero, an additional operation is performed:

metric = metric * [k5/(reliability + k4)]

Bandwidth is inverse minimum bandwidth of the path in BPS scaled by a factor of 2.56 * 1012. The range is from a 1200-bps line to 10 terabits per second.

Delay is in units of 10 microseconds. The range of delay is from 10 microseconds to 168 seconds. A delay of all ones indicates that the network is unreachable.

The delay parameter is stored in a 32-bit field, in increments of 39.1 nanoseconds. The range of delay is from 1 (39.1 nanoseconds) to hexadecimal FFFFFFFF (decimal 4,294,967,040 nanoseconds). A delay of all ones (that is, a delay of hexadecimal FFFFFFFF) indicates that the network is unreachable.

Troubleshooting EIGRP

Cisco has set of flow charts which I have personally found the best way to troubleshoot EIGRP.

Summarising EIGRP routes:-

We can use following command to summarise EIGRP routes. Under interface config mode, issue following command

ip summary address eigrp x.x.x.x subnet-mask.

EIGRP or OSPF

EIGRP has a faster convergence time as compare to OSPF. Administrative distance for internal EIGRP is 90 and that of external EIGRP is 170.

OSPF is still widely used dynamic protocol as compare to EIGRP, this is because EIGRP is cisco properietory and thus runs only on cisco routers where as OSPF is vendor independent routing protocol.

 

 

Advertisements

30 Comments »

  1. your article was really good. i appreciate your knowledge

    Comment by Karthi — November 11, 2006 @ 6:52 pm

  2. Nice Article, Expecting more documents from your side.
    regards,
    Mayank Nauni

    Comment by Mayank Nauni — January 2, 2007 @ 10:48 am

  3. Good and consice document.

    Comment by Sandeep — February 12, 2007 @ 12:42 pm

  4. Excellent Document to get a brief idea

    Comment by Sumeet M.Kale — April 12, 2007 @ 6:49 pm

  5. Excellent document, helpful for understanding the real use of Eigrp routing Protocol.

    Comment by Kiran Khairnar — January 18, 2008 @ 9:53 pm

  6. Dude ..ur efforts r really worth appriciating…we would surely love to learn more from u…

    Comment by Saurabh Joshi — January 25, 2008 @ 1:24 pm

  7. Its funtastic.

    Comment by Bahubali — February 6, 2008 @ 11:25 am

  8. Great, will help to compare OSPF and EIGRP. If can upload more topics on it.
    TQ

    Comment by Abdul Gafur Lodi — February 25, 2008 @ 6:45 am

  9. EIGRP is the best IGP???!?

    Be careful when you are making such a comment.

    OSPF & EIGRP have their own benefits and drawbacks, and I wouldn’t say EIGRP is the best.

    Listen to what Jeff Doyle says:
    http://www.networkworld.com/community/node/16276

    Comment by David — April 21, 2008 @ 7:12 am

  10. I’d agree with the previous poster that EIGRP may not be the best, and is certainly not the best in many applications. IS-IS is generally considered the best for large service providers needing speed and scalability. OSPF is generally considered best for heterogeneous enterprise networks and large enterprises. EIGRP may be best for small and medium enterprises running all Cisco, and not considering having to change (though IGP migrations are not too difficult if planned for well).

    Comment by Mike — May 27, 2008 @ 12:05 am

  11. 1. What are K1, K2, K3, K4 and K5?
    2. What is the purpose of using K Values in calculating metric?
    3. According to Documents I came to know that K1 = Bandwidth, K2=Load, K3=Delay,
    K4=Reliability and K5=MTU. If this is ture Why Cisco is recommending to multiply
    bandwidth with bandwidth and Delay with Delay?
    4. Why K1 to K5 values can be incremental from 1 to 255. Using (metric weights tos k1 k2
    k3 k4 k5)

    Comment by Venkata S. Dendukuri — April 21, 2009 @ 8:44 am

    • I have just made a related comment about the K constant descriptions which I think are misleading.

      Comment by Bill Laing — March 2, 2010 @ 5:30 am

  12. Just a note:

    No mtu in metric formula

    metric = [K1 * bandwidth + (K2 * bandwidth) / (256 − load) + K3 * delay] * [K5 / (reliability + K4)]

    the k values are k1=1, k2=0, k3=1, k4=0, k5=0

    so if you reduce the formula metric = bw + delay where bw is the lowest int bandwidth in the path but scaled or modified

    bandwidth = (10000000/bandwidth(i)) * 256

    delay is all the delays added along the path.

    source
    http://www.cisco.com/en/US/tech/tk365/technologies_white_paper09186a0080094cb7.shtml#eigrpmetrics

    Comment by lcappelli — July 20, 2009 @ 3:16 am

    • Icappelli,

      K1=Bandwidth
      k2=load
      k3=delay
      k4=reliability
      k5= MTU

      You can test your self.

      Comment by ciscotips — August 11, 2009 @ 2:53 pm

  13. Thank you for this tips

    Personal note:

    Please be respectful, these tips are for free and if you don’t entirely agree with them, say it calmly with respect.

    You can find the original documentation from Cisco on cisco.com

    Thank you

    Comment by Y Meziane — September 19, 2009 @ 1:32 pm

  14. the best doc i have read till now …regarding any protocol as it gives you everything…..what i needed for ccna .

    Comment by nitin — October 16, 2009 @ 11:28 pm

  15. Where do the EIGRP K constants acquire their descriptions from?

    I often see K1=Bandwidth, K2=Load, K3=Delay, K4=Reliability, K5=MTU in various articles but I have never seen it in any Cisco documentation.

    In what respect is K5 the MTU constant for instance? MTU is never used in the EIGRP metric calculation but K5 can be so how does K5 affect the value of MTU ?

    My contention is that the K values are just constants used to adjustthe metric calculation (if desired) and to give them descriptions is misleading and unneccessary.

    Bill Laing

    Comment by Bill Laing — March 2, 2010 @ 5:26 am

  16. many thanks for creating this site with documents

    Comment by varun — August 15, 2010 @ 7:22 pm

  17. Hi,
    i am pradeep singh

    Comment by Pradeep singh — September 1, 2010 @ 8:54 am

  18. Thanks for the good articles.
    hopefully useful.

    Regards,
    Grosir Sepatu Safety
    grosirsepatusafety.wordpress.com

    Comment by Grosir Sepatu Safety — September 14, 2010 @ 10:19 am

  19. […] CCIE notes for EIGRP July 2006 20 comments Possibly related posts: (automatically generated)#1 Page and Top 3 Posts In 2009 Leave a Comment LikeBe the first to like this post.Leave a Comment » […]

    Pingback by 2010 in review « Cisco Tips & Tricks — January 2, 2011 @ 6:55 am

  20. it’ nice

    Comment by IPS Malhi — April 1, 2011 @ 4:09 pm

  21. it’s nice

    Comment by IPS Malhi — April 1, 2011 @ 4:22 pm

  22. I was really confused on k values but thank you now i am clear.

    Comment by Lama — June 9, 2011 @ 5:50 am

  23. All the above trash, all K values are ranged from 0-255. IF you consider Bandwidth as K1 Value, you are scaling Bandwidth with Bandwidth, Ex: if the actual bandwidth of the link is 1544, answer should be 1544*1544.

    IS this correct. No not at all…. for your ref: check command: redistribute connected metric

    K Values are just integers to scale or manipulate your connected neighbor or to influence path selection – Traffic Engineering, they must agree between two directly connected neighbors.

    Metrics are different from K Values.

    if my comments found good please say Thanks, Don’t Mislead Network Engineers

    Venkata S. Dendukuri
    CCNA, CCNA-Route, CCIP BGP&MPLS, MCSE on Wndows Server 2003

    Comment by Venkata Dendukuri — October 24, 2011 @ 12:24 pm

  24. One more note:
    K1=K3=1, K2=K4=K5=0, is defined because in a worst case scenario, load, reliability and MTU are constantly changing, your EIGRP’s DUAL always calculating successor and feasible successor.

    As i mentioned in earlier post K-Values are used only to influence the path it is just EIGRP traffic engineering.

    Authors are copying each other, because they don’t have time to think.

    Again, Don’t mislead network engineers

    Comment by Venkata Dendukuri — October 24, 2011 @ 12:34 pm

  25. just bumped into this site from googling, so will keep this brief, …

    >This fancy name is being used by cisco as a kind of marketing tool.
    oops, marketing tool? (start here: see Dijkstra and Scholten on diffusing computations)

    >It has features of both distance vector & link state protocol
    care to entertain us by elaborating the above? 🙂

    > Feasible successor is nothing but a backup route.
    hmmm, hint: how does unequal load balancing under EIGRP process work?

    >EIGRP route metrics
    as an engineer you need to be more specific, maybe if you had worded this as metric weights

    > ..we can simply go under interface configuration mode and change delay
    be aware that this needs to be done across the whole EIGRP domain, else = “%DUAL-5-NBRCHANGE: EIGRP-IPv4 100: Neighbor 1.1.8.8 (Serial1/0) is down: K-value mismatch”

    >ip summary address eigrp x.x.x.x subnet-mask
    forgot something? (hint: AS)

    Comment by ABC — December 5, 2011 @ 11:06 pm

  26. EIGRP Summary: The key words “summary” and “address” in the command are connected by a hyphen “-”
    Command should be: “ip summary-address eigrp AS addresss mask” instead of “ip summary address eigrp AS addresss mask”

    Comment by realmatrix — August 21, 2012 @ 11:30 am

  27. Reblogged this on staniislaus.

    Comment by mrsugbome — October 26, 2014 @ 6:27 pm


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: