Superscope Question

rewindrewind Member Posts: 40 ■■□□□□□□□□
I have a superscope with two client scopes on a DHCP server. Neither client scope is full (there are available IPs to lease).

On a remote segment there is a relay agent configured. The remote segments default gateway has two IPs, one in each subnet. When a new client broadcasts a Discover on the remote network, and the relay agent forwards it, how does the DHCP server know which client scope to Offer a lease from??

Any thoughts or info is appreciated!

Comments

  • sprkymrksprkymrk Member Posts: 4,884 ■■■□□□□□□□
    The IP Address of the relay agent is used to determine the scope used. The relay agent is bound to the interface that is facing the clients it relays for. Let me know if I missed what you were asking.

    HTH. :)
    All things are possible, only believe.
  • rewindrewind Member Posts: 40 ■■□□□□□□□□
    sprkymrk wrote:
    The IP Address of the relay agent is used to determine the scope used. The relay agent is bound to the interface that is facing the clients it relays for. Let me know if I missed what you were asking.

    HTH. :)

    Thanks for replying :) I actually read the long post between you and royal from back in August on the same subject but still have a little confusion.

    I understand that in a normal scope the DHCP server will use the IP of the relay to determine which scope to offer a lease from. Since there are multiple child scopes in a superscope, but the relay agent can only have one ip assigned to it, how does the DHCP server know from which client scope to offer a lease? Does it alternate from one scope to the next, or does it just offer from a single child scope until it is full, then starts offering fromt the next?
  • sprkymrksprkymrk Member Posts: 4,884 ■■■□□□□□□□
    Oh, okay - I read your original question too fast.
    I have a superscope with two client scopes on a DHCP server. Neither client scope is full (there are available IPs to lease).

    On a remote segment there is a relay agent configured. The remote segments default gateway has two IPs, one in each subnet.

    Okay, as far as I can tell - the clients, even though they have 2 default gateways assigned, will only use the FIRST default gateway listed unless it is down. So they would then be assigned an IP based on the relay agent used for that gateway's network address.

    This would be a pretty easy and interesting experiemnt to set up by doing a release/renew on a client from each scope while disabling one default gateway at a time and seeing what IP it gets.
    All things are possible, only believe.
  • royalroyal Member Posts: 3,353
    Read this. It clarifies the topic that Mark and I were talking about.
    http://www.shudnow.net/2007/11/20/dhcp-scope-vs-superscope/
    “For success, attitude is equally as important as ability.” - Harry F. Banks
Sign In or Register to comment.