Upgrade from 3.0.2 to 3.0.4 and some strange changes
Hall J D (ISeLS)
jdhall at glam.ac.uk
Thu Oct 12 09:15:10 UTC 2006
Hello all,
I've just upgraded from 3.0.2 to 3.0.4 on my pair of DHCP servers and
I've noticed one or two strange things since the upgrade that I hope
someone can explain for me.
All though we are running failover almost all of the clients get their
addresses statically via host statements. Previously if a client tried
to request an address, and we hadn't defined a host statement for it,
we'd get a "network vlan138: no free leases" message logged on both
servers. Since the upgrade we now get a "peer holds all free leases"
message on both servers instead!
Here's a config snipit for one of our subnets:-
#
# Class used to identify rouge RAS servers and give them a
# non-routable 192.168.x.x address to shut them up. We can then use the
host
# name they record in the leases file to track them down.
#
class "ras-jail" {
match if substring (option dhcp-client-identifier, 1, 3) =
"RAS";
}
shared-network vlan138 {
subnet 193.63.138.0 netmask 255.255.255.0 {
option routers 193.63.138.254;
}
subnet 193.63.158.0 netmask 255.255.255.0 {
option routers 193.63.158.254;
}
subnet 193.63.159.0 netmask 255.255.255.0 {
option routers 193.63.159.254;
}
subnet 192.168.138.0 netmask 255.255.255.0 {
pool {
failover peer "glam";
deny dynamic bootp clients;
range 192.168.138.1 192.168.138.253;
allow members of "ras-jail";
}
}
}
Does anyone know why the message has changed?
Something else I noticed is some strange dates in the failover peer
state in my leases file. On my primary server the state is:-
failover peer "glam" state {
my state normal at 4 2006/10/12 08:46:01;
partner state normal at 4 2001/10/18 08:24:58;
}
While on the secondary it's:-
failover peer "glam" state {
my state normal at 4 2006/10/12 08:45:01;
partner state normal at 4 2006/10/12 06:50:42;
mclt 3600;
}
Any idea why is the partner state date on the primary server 2001?
Thanks,
Jonathan
More information about the dhcp-users
mailing list