Geographic distant dhcp servers and failover

[Leith Patrick-LTHP001]

Hi, 
 
Thanks for the update. I have yet another question. 
 
I have proved that a v4 lease db is backward compatible to v3.1.3. But
due to issues relating to Redback switch/routers and existing
established tunnel resource allocation the system has to keep lease
times to a minimum (15 minutes) as Redback does not have the
functionality to release resources. 
 
The current dhcp server is dealing with a client base of 50000. This is
expected to double and I am concerned that the server will not be able
to cope. 
 
The plan is to split the IP pools on two geographically distant servers.
But this will mean there is no redundancy. If one server were to fail is
there a recommended method to synchronise the failed dhcp's lease db and
enable it's IP pool on the remaining dhcp server? IP split but with
failover. 
 
Bets regards, 
 
Patrick. 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/dhcp-users/attachments/20091111/cc65424d/attachment.html>