- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Fixed address and DHCP Failover pairs
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
05-29-2018 11:55 AM
Hi,
I'm coming from a QIP installation which has the M-DHCP (Manual) construct which, from what I understand, is similar to a "fixed address" in Infoblox. The question I have is around DHCP failover pairs which are placed on a DHCP range template, but not on a DHCP fixed address or template. Say a client is configured with a fixed address in infoblox; the client will initially try to request an address via DISCOVER. Why wouldn't a faiolver pair be appropriate in this case if the primary DHCP server is down? Thanks.
Brad
Solved! Go to Solution.

Re: Fixed address and DHCP Failover pairs
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
05-30-2018 06:13 PM
Hello Brad,
DHCP Failover (DHCPFO) allows an administrator to add redundancy for dynamic pools of DHCP addresses, but is not used for DHCP fixed addresses.
A DHCP fixed address is a "reserved" IP address that is usually mapped to a specific MAC address. When a client, on a particular subnet, requests an IP address, a DHCP server will do one of a couple of things:
- If the network is not defined, or there are no addresses to hand out, the server does nothing.
- If the server has a pool of addresses for generic use, it can hand out a "random" IP address.
- If the server has a "dedicated" IP address for the client, the server will hand out that IP address.
Rule number 3 will actually take precedence so that even if there is a generic pool of addresses available, the fixed-address will always be used.
When you map a DHCP network to a member (or system if not using grid), all fixed-address objects defined for that network are written to that member. So, if using DHCP Failover, you would have a network that was mapped to (at least) two members. All fixed-addresses would also get written to both members. This would mean that either member could hand out the IP address (or both at the same time!) but this does not break anything as they would both be handing out the same IP address to the same exact client. This allows for redundancy for the fixed-address as even if one of the servers goes down, the others will be able to answer the request.
Hope this answers your question.
Regards,
Manu M