ip classless and auto-summary

Discussion in 'Cisco' started by jh3ang, Aug 4, 2006.

  1. jh3ang

    jh3ang Guest

    how are you everyone ?

    I'm just a little bit confused about these two commands:
    ip classless and auto-summary.

    It sounds like both commands have the same function..
    Please tell me what the differnece is..

    I have read a lot about this commands, but I couldn't find a point -.-;

    when auto-summary is used ?
    if i use no auto-summarry, do i still have to concern about ip classess
    ?
     
    jh3ang, Aug 4, 2006
    #1
    1. Advertisements

  2. jh3ang

    Anonymous Guest

    "ip classless" should always be enabled. Back in the old days, the first
    octet of the IP address determined its "class" 1 through 127 were class A,
    128 through 191, were class B, 192 to 223 were Class C, 224 to 239 were
    Class D, and 240 to 255 were Class E. The class determined the network
    mask. Class A networks had 255.0.0.0 mask, class B, 255.255.0.0, class C
    255.255.255.0, class D is reserved for multicast, and class E was reserved
    This idea went out a long time ago, and now class does not have any meaning
    except for Class D, which is reserved for multicast, and Class E, which is
    still reserved.

    auto-summary is a similar concept, but is used for routing, where the
    routing protocol would "auto-summarize" the routes using the above obsolete
    mechanism. You should also enable "no auto-summary" as well.

    "classfull" addresses and "auto-summary" were the original IOS defaults way
    back when we had a "classfull" Internet. Today that is not the case, but
    Cisco very, very, very rarely ever changes IOS default settings.

    Scott
     
    Anonymous, Aug 5, 2006
    #2
    1. Advertisements

  3. jh3ang

    Merv Guest

    Merv, Aug 5, 2006
    #3
  4. "ip classless" is used when the router is searching its routing table to
    determine how to forward a packet. Without this option, if it has a
    route for any subnet of a classic network, and it's searching for a
    route within that network, it will not use any aggregates outside the
    network. E.g. if you have routes:

    172.15.10.0/24
    172.14.0.0/15

    and you're trying to send to 172.15.20.1, it will report "No route",
    even though 172.14.0.0/15 matches. In effect, the 172.15.10.0/24 route
    causes it to create a virtual null route for the 172.15 class B network.

    Since a default route is just the ultimate aggregate, the most common
    cause of leaving out "ip classless" is that packets for unknown subnets
    are dropped rather than being sent to the default gateway.

    "auto-summary" is use by routing protocols. It causes the router to
    *advertise* this virtual route when it has routes for the subnets.

    If you have both auto-summary and ip classless, it's easy to create
    routing loops. If the router advertises the class B to its default
    gateway, the gateway will send packets for any subnet to this router.
    But if the router doesn't have a specific route for that subnet it will
    send the packets back to the default gateway, and they'll go back and
    forth.

    What I recommend instead of auto-summary is that you create an explicit
    Null0 route for the class B if you want to advertise it.
     
    Barry Margolin, Aug 5, 2006
    #4
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.