EIGRP metrics...Max hopcount 224...what?

NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
Okay so EIGRP uses Bandwidth and delay by default to calculate the metric. Thats great.
Also EIGRP sends a full routing table once to a neighbor and then only sends changes to neighbors after convergence. Thats great too. Now this may be beyond the scope of CCNA, but whats up with a Maximum hopcount of 224? I know load,reliability and maximum transmission unit can play a role in EIGRP behaviour. I also realize that IGRP probably uses 255 for route poisoning seeing as it uses bandwidth and delay as well. Now EIGRP doesn't use distance vector loop prevention mechanisms because it stores topological data and it does not have a default hop count because it is a hybrid.

So how and when does max. hop count come into play with EIGRP?
There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!

Comments

  • flares2flares2 Member Posts: 79 ■■□□□□□□□□
    EIGRP does have a maximum hop count. By default the max hop count for EIGRP is 100. This can be changed anywhere from 1 all the way to 255 using the command (config-router)#metric maximum-hops <1-255>
    As far as I know, the hop count does not play a role in calculating the best path. It's only purpose is to prevent routing loops.
    Techexams.net - Job security for one more day.
  • dpocorobadpocoroba Member Posts: 12 ■□□□□□□□□□
    EIGRPs max hop count should be 255

    the range for hop count in the header is 0x00 to 0xff ( FF = 255 )

    Its basicly the same concept as OSPF really having a max hop count of 65535 assuming all the links are set to a cost of 1.


    DP
    "Knowledge is contagious, infect"
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    flares2 wrote:
    EIGRP does have a maximum hop count. By default the max hop count for EIGRP is 100. This can be changed anywhere from 1 all the way to 255 using the command (config-router)#metric maximum-hops <1-255>
    As far as I know, the hop count does not play a role in calculating the best path. It's only purpose is to prevent routing loops.

    Actually 1-255 is for IGRP default is 100

    Max. hop count for EIGRP is 224 with no default. READ THE TECHNOTES
    EIGRP prevents routing loops by using a topology database, DUAL, and RTP.

    So why would max hop 224 be in the books? It has to be used somehow or it would never have been written.
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • flares2flares2 Member Posts: 79 ■■□□□□□□□□
    Netstudent wrote:
    Actually 1-255 is for IGRP default is 100

    Max. hop count for EIGRP is 224 with no default. READ THE TECHNOTES

    EIGRP prevents routing loops by using a topology database, DUAL, and RTP.

    So why would max hop 224 be in the books? It has to be used somehow or it would never have been written.

    Sybex 5th Edition page 298 Paragraph 1:
    "In addition, EIGRP has a maximum hop count of 100, but can be set up
    to 255. Chances are you wouldn’t want to ever change this, but if you did, here is how you
    would do it:
    Pod1R1(config)#router eigrp 10
    Pod1R1(config-router)#metric maximum-hops ?
    <1-255> Hop count
    As you can see from the router output above, EIGRP can be set to a maximum of 255 hops,
    and even though it doesn’t use hop count in the path metric calculation, it still uses the maximum
    hop count to limit the scope of the AS."

    Technotes may say 224, and nothing against technotes, but Todd Lammle says 100 with a max of 255.
    Techexams.net - Job security for one more day.
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    HMM. thats interesting because maxhop of 224 is all over the internet as well.
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • CucumberCucumber Member Posts: 192
    whats up with a Maximum hopcount of 224?
    

    It s because of the queries and replies.
    When EIGRP losses all non-loop routes for a destination in the topology table it will send queries for it, the queries then are flooded. The more hops, the more flood. Also, a larger hop count could mean the replies for those queries may get lost in the network, they must get back in 3 minutes or else EIGRP gets "stuck in active mode".
    Actually, Cisco recomends having a maximum of 7 hops in an EIGRP network.
    Now EIGRP doesn't use distance vector loop prevention mechanisms because it stores topological data and it does not have a default hop count because it is a hybrid. 
    

    Not really, EIGRP does use Split Horizon as a loop prevention mechanism.
    You can even run into problems using a Hub And SPoke frame relay topology when using EIGRP because of the Split Horizon thing, unless you use subinterfaces.

    EIGRP is actually an Advanced Distance Vector protocol, which means it is like any distance vector protocol but just with a few improvements, its topology table is far from being as
    complete as OSPF's or ISIS's. Calling it hybrid is an over statement in my opinion.
    I hate pandas
  • flares2flares2 Member Posts: 79 ■■□□□□□□□□
    Well Netstudent, to further confuse and/or disprove both of us, this came from the below link:

    http://www.cisco.com/warp/public/103/eigrp-toc.html

    The maximum number of hops that EIGRP will accept is 100 by default, although the maximum can be configured to 220 with metric maximum hops.
    Techexams.net - Job security for one more day.
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    OKay now I understand why and how a hopcount is used. And I just read about split-horizon and EIGRP in my BSCI book after my first post. I also didn't find anything on hop count with EIGRP in BSCI. So now what is it? 224 or 255? I looked in my sybex book and Lammle said 255 just like Flares mentioned. And how can you argue with the IOS output? Having said that, why is 224 strewn all over the internet and technotes. Has there been an IOS version update?
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    flares2 wrote:
    Well Netstudent, to further confuse and/or disprove both of us, this came from the below link:

    http://www.cisco.com/warp/public/103/eigrp-toc.html

    The maximum number of hops that EIGRP will accept is 100 by default, although the maximum can be configured to 220 with metric maximum hops.

    ya I read some of that article as well. I'm wondering if maybe there is a difference betwen external EIGRP and internal EIGRP
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • CucumberCucumber Member Posts: 192
    This is what the Cisco IOS IP Routing Protocols Command Reference Release 12.2SR says:

    metric maximum-hops
    To have the IP routing software advertise as unreachable those routes with a hop count higher than is specified by the command (Enhanced Interior Gateway Routing Protocol [EIGRP] only), use the metric maximum-hops command in router configuration mode. To reset the value to the default, use the no form of this command.

    Maximum hop count (in decimal). The default value is 100 hops; the
    maximum number of hops that can be specified is 255.
    I hate pandas
  • dpocorobadpocoroba Member Posts: 12 ■□□□□□□□□□
    You would be surprised how many things are mis-documented in the Cisco site.

    http://www.cisco.com/univercd/cc/td/doc/product/software/ios124/124cr/hirp_r/rte_eih.htm#wp1000064
    "Knowledge is contagious, infect"
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    Okay then I'm putting 255 in my brain and it's there to stay.
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    Cucumber wrote:
    whats up with a Maximum hopcount of 224?
    

    It s because of the queries and replies.
    When EIGRP losses all non-loop routes for a destination in the topology table it will send queries for it, the queries then are flooded. The more hops, the more flood. Also, a larger hop count could mean the replies for those queries may get lost in the network, they must get back in 3 minutes or else EIGRP gets "stuck in active mode".
    Actually, Cisco recomends having a maximum of 7 hops in an EIGRP network.
    Now EIGRP doesn't use distance vector loop prevention mechanisms because it stores topological data and it does not have a default hop count because it is a hybrid. 
    

    Not really, EIGRP does use Split Horizon as a loop prevention mechanism.
    You can even run into problems using a Hub And SPoke frame relay topology when using EIGRP because of the Split Horizon thing, unless you use subinterfaces.

    EIGRP is actually an Advanced Distance Vector protocol, which means it is like any distance vector protocol but just with a few improvements, its topology table is far from being as
    complete as OSPF's or ISIS's. Calling it hybrid is an over statement in my opinion.

    So cisco recommends no more than 7 hops to prevent SIA's. Does that mean you have to redistribute multiple Autonomous systems with each Autonomous system having no more than 7 hops? Theoretically according to Cisco?

    Great discussion, thanks for your input guys.... icon_wink.gif
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • CucumberCucumber Member Posts: 192
    You can use EIGRP stub routing and manual summarization to avoid SIA problems before thinking of breaking up the autonomous system.
    I hate pandas
  • NetstudentNetstudent Member Posts: 1,693 ■■■□□□□□□□
    oh ok..thats a little beyond me right now...I'll be there soon though.
    There is no place like 127.0.0.1 BUT 209.62.5.3 is my 127.0.0.1 away from 127.0.0.1!
  • CucumberCucumber Member Posts: 192
    I think SIAs and stub routing is beyond CCNA, anyway I think I diverted the point of your question, I was just busy trying to make a point about why hop count could matter to EIGRP, hehe sorry.

    As people said before me, the maximum hop-count parameter is used to avoid counting to infinity (avoid loops) in EIGRP.
    I hate pandas
Sign In or Register to comment.