Well, I seem to be getting somewhere, although where exactly is open to question.
I did this. I put the virtual interface address 192.168.0.1 back onto eth1 of the gateway host and restarted the network services. The ifcfg file looked like this:
BOOTPROTO=none BROADCAST=192.168.255.255 DEVICE=eth1:192 IPADDR=192.168.0.1 IPV6INIT=no MTU="" NAME="LAN - Non-routable" NETMASK=255.255.0.0 NETWORK=192.168.0.0 ONBOOT=yes ONPARENT=yes
After the restart ip addr showed this:
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:25:90:60:11:8d brd ff:ff:ff:ff:ff:ff inet aaa.bbb.ccc.1/24 brd aaa.bbb.ccc.255 scope global eth1 inet 192.168.0.1/24 brd 192.168.255.255 scope global eth1:192 inet6 fe80::225:90ff:fe60:118d/64 scope link valid_lft forever preferred_lft forever
Note the cidr suffix on 192.168.0.1 = 24
That is not what I expected. Restarting with the same config did not change the initially observed outcome.
SO, I edited ifcfg-eth1:192 and added exactly one line:
PREFIX="16"
and restarted the network. ip addr now shows this:
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:25:90:60:11:8d brd ff:ff:ff:ff:ff:ff inet aaa.bbb.ccc.1/24 brd aaa.bbb.ccc.255 scope global eth1 inet 192.168.0.1/16 brd 192.168.255.255 scope global eth1:192 inet6 fe80::225:90ff:fe60:118d/64 scope link valid_lft forever preferred_lft forever
Note that the cidr suffix is now 16.
Now, when I try and ping an address on the 192.168 netblock from host C I see this:
# ping 192.168.209.43 PING 192.168.209.43 (192.168.209.43) 56(84) bytes of data.
From 216.185.71.1: icmp_seq=2 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=3 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=4 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=5 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=6 Redirect Host(New nexthop: 192.168.209.43)
My question now is how do I get to 192.168.209.43?
On Thu, Sep 6, 2012 at 3:54 PM, James B. Byrne byrnejb@harte-lyne.ca wrote:
I did this. I put the virtual interface address 192.168.0.1 back onto eth1 of the gateway host and restarted the network services. The ifcfg file looked like this:
BOOTPROTO=none BROADCAST=192.168.255.255 DEVICE=eth1:192 IPADDR=192.168.0.1 IPV6INIT=no MTU="" NAME="LAN - Non-routable" NETMASK=255.255.0.0 NETWORK=192.168.0.0 ONBOOT=yes ONPARENT=yes
After the restart ip addr showed this:
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:25:90:60:11:8d brd ff:ff:ff:ff:ff:ff inet aaa.bbb.ccc.1/24 brd aaa.bbb.ccc.255 scope global eth1 inet 192.168.0.1/24 brd 192.168.255.255 scope global eth1:192 inet6 fe80::225:90ff:fe60:118d/64 scope link valid_lft forever preferred_lft forever
Note the cidr suffix on 192.168.0.1 = 24
That is not what I expected. Restarting with the same config did not change the initially observed outcome.
SO, I edited ifcfg-eth1:192 and added exactly one line:
PREFIX="16"
and restarted the network. ip addr now shows this:
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:25:90:60:11:8d brd ff:ff:ff:ff:ff:ff inet aaa.bbb.ccc.1/24 brd aaa.bbb.ccc.255 scope global eth1 inet 192.168.0.1/16 brd 192.168.255.255 scope global eth1:192 inet6 fe80::225:90ff:fe60:118d/64 scope link valid_lft forever preferred_lft forever
Note that the cidr suffix is now 16.
I thought it would figure that out from the NETMASK, but OK....
Now, when I try and ping an address on the 192.168 netblock from host C I see this:
# ping 192.168.209.43 PING 192.168.209.43 (192.168.209.43) 56(84) bytes of data.
From 216.185.71.1: icmp_seq=2 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=3 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=4 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=5 Redirect Host(New nexthop: 192.168.209.43) From 216.185.71.1: icmp_seq=6 Redirect Host(New nexthop: 192.168.209.43)
My question now is how do I get to 192.168.209.43?
This is your router telling the source box that it can send directly to the destination (which it knows because netmasks really are supposed to be global for the subnet and routers don't like to route back the inbound interface). However, it should also have routed the packet.
On 09/06/2012 11:11 PM, Les Mikesell wrote:
On Thu, Sep 6, 2012 at 3:54 PM, James B. Byrne byrnejb@harte-lyne.ca wrote:
I did this. I put the virtual interface address 192.168.0.1 back onto eth1 of the gateway host and restarted the network services. The ifcfg file looked like this:
BOOTPROTO=none BROADCAST=192.168.255.255 DEVICE=eth1:192 IPADDR=192.168.0.1 IPV6INIT=no MTU="" NAME="LAN - Non-routable" NETMASK=255.255.0.0 NETWORK=192.168.0.0 ONBOOT=yes ONPARENT=yes
After the restart ip addr showed this:
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:25:90:60:11:8d brd ff:ff:ff:ff:ff:ff inet aaa.bbb.ccc.1/24 brd aaa.bbb.ccc.255 scope global eth1 inet 192.168.0.1/24 brd 192.168.255.255 scope global eth1:192 inet6 fe80::225:90ff:fe60:118d/64 scope link valid_lft forever preferred_lft forever
Note the cidr suffix on 192.168.0.1 = 24
That is not what I expected. Restarting with the same config did not change the initially observed outcome.
SO, I edited ifcfg-eth1:192 and added exactly one line:
PREFIX="16"
and restarted the network. ip addr now shows this:
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:25:90:60:11:8d brd ff:ff:ff:ff:ff:ff inet aaa.bbb.ccc.1/24 brd aaa.bbb.ccc.255 scope global eth1 inet 192.168.0.1/16 brd 192.168.255.255 scope global eth1:192 inet6 fe80::225:90ff:fe60:118d/64 scope link valid_lft forever preferred_lft forever
Note that the cidr suffix is now 16.
I thought it would figure that out from the NETMASK, but OK....
It does.
The question is what does the config file for eth1 look like because when you bring up an alias interface first the config file for the parent interface is read and then those values are overwritten by the values in the alias config file. So it might be the case that there is a PREFIX=24 definition in the eth1 file and none in the eth1:192 file which so in the end PREFIX=24 would be used for the alias interface.
Regards, Dennis