DHCPD , dhcp relays on a large network
Alexandru Coseru
alexandru.coseru at totaltelecom.ro
Fri Nov 10 14:41:35 UTC 2006
Quote: "> This doesn't sound right, the server will NOT fail to start
because a remote subnet is not defined - all that would happen is you would
get runtime errors in response to requests for unknown subnets."
It is the other way around.
Srv1 has:
[root at bv ~]# ifconfig
eth0 Link encap:Ethernet HWaddr 00:0C:46:B3:4F:2F
inet addr:10.1.1.1 Bcast:10.1.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:35 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 b) TX bytes:5384 (5.2 KiB)
Interrupt:17 Base address:0x2c00
[root at bv ~]#
cat /etc/dhcpd.conf:
ddns-update-style ad-hoc;
authoritative;
subnet 192.168.1.0 netmask 255.255.255.0{
option routers 192.168.1.1;
option subnet-mask 255.255.255.0;
option nis-domain "domain.org";
option domain-name "domain.org";
option domain-name-servers 10.1.1.1;
default-lease-time 21600;
max-lease-time 43200;
host phonehuawei {
hardware ethernet 88:88:88:88:88:88;
fixed-address 192.168.1.2;
}
host universalphone{
hardware ethernet 00:13:49:53:C6:3B;
fixed-address 192.168.1.3;
}
host accountingserver{
hardware ethernet 00:43:29:12:12:FA;
fixed-address 192.168.1.4;
}
}
The Layer3 switch has DHCPD relay agent on it... and 2 IP addresses:
10.1.1.2 on management vlan (Gbit port) and 192.168.1.1 on clients vlan
The client voip phone should broadcast a request for an ip address , the
L3 SW A is forwarding that request to srv1 (I can see it using tcpdump).
If i'm not declaring the subnet =>runtime errors , request for unknow
subnets.
Using the dhcpd config described earlier , i have:
"No subnet declaration for 192.168.1.0".
Regards
Alex
----- Original Message -----
From: "Simon Hobson" <dhcp1 at thehobsons.co.uk>
To: <dhcp-users at isc.org>
Sent: Friday, November 10, 2006 4:19 PM
Subject: Re: DHCPD , dhcp relays on a large network
> Alexandru Coseru wrote:
>
>>The ideea is that each switch has a subnet attached to it..
>>
>>On the backbone , I have the dhcpd server..
>>
>>The dhcpd server is not directly attached to the subnets , it has to go
>>through a layer3 route to reach them..
>>
>>
>> 10.1.1.0/24
>> 192.168.1.0/24
>>srv1 ---------------- L3 SW A -------------------------
>>subnet1
>
> So far it's looking like a normal routed network.
>
>>The L3 SW A has DHCPD relay options enabled , so the DHCPD requests
>>arrives on srv1.
>>But srv1 has no knowledge of class 192.168.1.0/24 , and the server
>>cannot be started (Error: No subnet declaration for 192.168.1.0/24).
>>The server has only one ethernet NIC , with ip 10.1.1.1 on it..
>
> This doesn't sound right, the server will NOT fail to start because a
> remote subnet is not defined - all that would happen is you would get
> runtime errors in response to requests for unknown subnets.
>
>>Can you elaborate more on shared networks ?
>
> It's a common source of confusion, but I don't think you have one. A
> shared subnet is where you have two different IP subnets on the same
> network segment (or if you are being pedantic on the same broadcast
> domain). In Linux for example, you can add additional IP addresses to
> an ethernet port, eg :
>
> eth0 - a.b.c.d/27
> eth0:0 - 192.168.0.1/24
> eth0:1 - 172.16.0.0/12
>
> In this topology, a device attached to this ethernet network could
> have an address in any one or more of the three subnets. This is
> indicated to dhcpd via the shared network construct as described in
> the man pages.
>
>
> Coming back to your problem, can you post the results of 'ifconfig'
> and the contents of your dhcpd.con file ?
>
> Simon
>
>
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.409 / Virus Database: 268.14.1/527 - Release Date: 11/9/2006
>
>
More information about the dhcp-users
mailing list