χ Fri, 07.06.2002, Χ 19:27, Christophe Prevotaux ΞΑΠΙΣΑΜ: > try to setup this (it won't work) your example does not work
As I already said you need to remove interface route here (or apply changes sugessted by Iasen Kostoff <[EMAIL PROTECTED]>) > ifconfig fxp0 inet 192.168.1.1/24 up > ifconfig fxp0 inet 192.168.1.1/32 alias route delete 192.168.1.1 > ifconfig fxp0 inet 192.168.1.1/32 alias <--- here it will not work anymore With route line it will work > this is the same thing with vlan interfaces under FreeBSD > you can't have more than 1 interface with the same IP/NETMASK > > ifconfig fxp0 inet 192.168.1.1/24 up > ifconfig create vlan0 inet vlan 1 vlandev fxp0 192.168.1.1/32 up route delete 192.168.1.1 > ifconfig create vlan1 inet vlan 2 vlandev fxp0 192.168.1.1/32 up <-- this does not >work before each new vlan initialization you need remove interface route > let's say for the example sake , that you want only to have 1 vlan on the fxp0 > then you can setup the interface as shown above but you can't route anything thru it > > route add 192.168.1.2 -interface vlan0 > > will not route properly the packets thru vlan0 either because it is a /32 (AFAIK) It is not valid command, see: # route add 1.2.3.4 -interface fxp0 add host 1.2.3.4: gateway fxp0 # netstat -rn | fgrep 1.2.3.4 1.2.3.4 00:00:00:00:00:00 UHLS 0 0 fxp0 # It is invalid route, -interface command add route direct reachable via interface (ppp interface for example) but there we have ethernet with indirect access (you need mac-resolve first) see: # route add -net 1.2.3.4/32 -cloning -iface fxp0 add net 1.2.3.4: gateway fxp0 # netstat -rn | fgrep 1.2.3.4 1.2.3.4/32 link#3 UCS 0 0 fxp0 # So, it is net-route and it will be cloned at request (arp-request will be sent) # ping -c1 1.2.3.4 PING 1.2.3.4 (1.2.3.4): 56 data bytes 64 bytes from 1.2.3.4: icmp_seq=0 ttl=64 time=1.799 ms --- 1.2.3.4 ping statistics --- 1 packets transmitted, 1 packets received, 0% packet loss round-trip min/avg/max/stddev = 1.799/1.799/1.799/0.000 ms # netstat -rn | fgrep 1.2.3.4 1.2.3.4 00:01:03:d6:9e:9f UHLW 0 2 fxp0 1192 => 1.2.3.4/32 link#3 UCS 1 0 fxp0 # you see at this stage you have cloning(C) netroute and cloned(W) hostroute to your NIC with 1.2.3.4 If you do not want arp-resolve you can specify MAC-address (link-layer address): # route add -host 4.3.2.1 -link aa:bb:cc:dd:ee:ff -ifp fxp0 add host 4.3.2.1: gateway aa:bb:cc:dd:ee:ff # netstat -rn | fgrep 4.3.2.1 4.3.2.1 aa:bb.cc.dd.ee.ff UGHS 0 0 fxp0 # Ruslan Ermilov <[EMAIL PROTECTED]> knows more about routes setup and may be correct me if I am mistaken. > and even if you could do it , then the problem is that you are limited to having > only one vlan per interface, which sucks BIG TIME. No, see above > under Linux you can have the equivalent of the following setup: > > 192.168.1.2/24 > 192.168.1.3/24 > 192.168.1.4/24 > > are machines behind the switch vlan 1 2 3 ports respectively > > ifconfig fxp0 inet 192.168.1.1/24 up > ifconfig create vlan0 inet vlan 1 vlandev fxp0 192.168.1.1/24 up > ifconfig create vlan1 inet vlan 2 vlandev fxp0 192.168.1.1/24 up > ifconfig create vlan2 inet vlan 3 vlandev fxp0 192.168.1.1/24 up > > etc... etc... > > and then you do the equivalent of > > route add 192.168.1.2 -interface vlan0 > route add 192.168.1.3 -interface vlan1 > route add 192.168.1.4 -interface vlan2 As I said -interface flag is invalid here, Linux or too smart or too stupid here. > and it works just fine, with no problems whatsoever. > > What I want to do is just this but under FreeBSD. > > My question is : How can I do it ? See above PS: As I already said, route manpage need some work, because it is very unclean, may be somebody who knows English well will do it ... -- Vladimir B. Grebenschikov [EMAIL PROTECTED], SWsoft, Inc. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-net" in the body of the message
