Rebooting router necessary?

[erpete]

I use OpenBSD as an experimental home router and there is a wifi external device attached to it as an access point. It works well so far. The only problem I have is when I switch off and on the wifi it has no access to internet then and I need to reboot the router. Wifi connects to router just fine but doesn't get connected to outer internet so I guess it's Packet Filter. Is there any way around it?

[fvgit]

Not sure if this helps, but you can restart your network without rebooting by running the netstart(8) script. It's what the system does when it boots.

See the manpage for further explanation. You can also perform a dry-run of netstart with the -n option. This will simply print all the commands netstart would run if invoked normally.

[J65nko]

As a start you could run these commands when you rebooted your OBSD pf router and WiFi clients can access the internet:

# (ifconfig -A ; netstat -rn -f inet ; cat /etc/resolv.conf) >working_log.txt

Then switch off and turn on the WiFi access point without rebooting your router to make sure your WiFi clients cannot access the Internet.

# (ifconfig -A ; netstat -rn -f inet ; cat /etc/resolv.conf) >notworking_log.txt

Now compare the the output of these 2 files:

# diff -u working_log.txt notworking_log.txt

Alternatively you also can do a sdiff(1) on these files.

[erpete]

Thank you, fvgit. I tried the netstart script before and it did not help. Didn't know about the -n option, though.

[erpete]

J65nko,

the relevant part (plus = notworking_log):

-192.168.2/24       192.168.2.1        UCn        3       18     -     4 ure1 
-192.168.2.1        d0:37:45:ec:fc:9d  UHLl       0       73     -     1 ure1 
-192.168.2.16       94:be:46:b6:24:a7  UHLc       1     1684     -     3 ure1 
-192.168.2.19       e4:c3:2a:65:c3:30  UHLc       0        3     -     3 ure1 
-192.168.2.20       e4:c3:2a:65:c3:30  UHLc       1       81     -     3 ure1 
+192.168.2/24       192.168.2.1        UCn        0        0     -     4 ure1 
+192.168.2.1        d0:37:45:ec:fc:9d  UHLl       0        0     -     1 ure1 
 192.168.2.255      192.168.2.1        UHb        0        0     -     1 ure1

Apparently, the wifi router and connected device (a tablet) are missing in the routing table. I just don't know why. The ifconfig and resolv.conf outputs are the same.

[J65nko]

I have a similar setup. An Alix 386 OpenBSD system with three NICs, one external, one for the internal wired Ethernet home LAN, and one for the WiFi router/access point.

On the Alix dhcpd has been configured to handle requests from both the internal wired LAN as well as from the WiFi clients. The following snippet is for the WiFi clients on the 10.7.7.0/24 NIC

     40 subnet 10.7.7.0  netmask 255.255.255.0 {
     41
     42         option routers 10.7.7.10;
     43         range 10.7.7.11  10.7.7.254;
     44
     45 }

Because the Alix does give out the adresses to the WiFi clients, it has no problems when the Wifi router/access point has been turned off/on.

Who gives out the addresses for your WiFi clients?Your WiFi router or your OpenBSD pf router?

[erpete]

In fact, there is subnet for LAN as well. The wifi part is:

subnet 192.168.2.0 netmask 255.255.255.0 {
        option domain-name-servers 192.168.2.1;
        option routers 192.168.2.1;
        range 192.168.2.10 192.168.2.254;
}

So OpenBSD router gives out the addresses. There is no such problem on LAN when connecting computers which is interesting.

[J65nko]

How is the WiFi access point connected to the OpenBSD pf router? How many NIC's does your OpenBSD pf box have? Three?
All three NICs should be on a different subnet.
What is the output of # ifconfig -A?
This is mine:

vr0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        lladdr 00:0d:b9:2a:bf:e0
        priority: 0
        groups: egress
        media: Ethernet autoselect (100baseTX full-duplex)
        status: active
        inet 192.168.2.3 netmask 0xffffff00 broadcast 192.168.2.255
vr1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        lladdr 00:0d:b9:2a:bf:e1
        priority: 0
        groups: internal
        media: Ethernet autoselect (100baseTX full-duplex)
        status: active
        inet 192.168.222.10 netmask 0xffffff00 broadcast 192.168.222.255
vr2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        lladdr 00:0d:b9:2a:bf:e2
        priority: 0
        groups: internal
        media: Ethernet autoselect (100baseTX full-duplex)
        status: active
        inet 10.7.7.10 netmask 0xffffff00 broadcast 10.7.7.255
        inet 10.7.7.11 netmask 0xffffffff

• vr0, the external interface is connected to my ISP router, and gets it IP from that router.
• vr1is an internal interface connected to an ethernet switch for my wired LAN clients.
• vr2is the second internal interface directly connected to my WiFi access point/router for the WiFi clients

As you can see they are all on different subnets.

[erpete]

Wifi is connected to external NIC (ure1, 192.168.2). LAN is connected to external NIC (ure0, 192.168.1). All via cable.

re0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        lladdr 90:1b:0e:56:ba:6c
        index 1 priority 0 llprio 3
        groups: egress
        media: Ethernet autoselect (1000baseT full-duplex,rxpause,txpause)
        status: active
        inet 192.168.8.110 netmask 0xffffff00 broadcast 192.168.8.255
...
ure0: flags=8c43<UP,BROADCAST,RUNNING,OACTIVE,SIMPLEX,MULTICAST> mtu 1500
        lladdr 7c:c2:c6:1c:6b:86
        index 4 priority 0 llprio 3
        media: Ethernet autoselect (100baseTX full-duplex,rxpause,txpause)
        status: active
        inet 192.168.1.1 netmask 0xffffff00 broadcast 192.168.1.255
ure1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        lladdr d0:37:45:ec:fc:9d
        index 5 priority 0 llprio 3
        media: Ethernet autoselect (none)
        status: no carrier
        inet 192.168.2.1 netmask 0xffffff00 broadcast 192.168.2.255

OBSD router is connected to ISP modem from internal NIC (re0) also via cable and receives IP address 192.168.8.110 from there (dhcp in hostname.re0). Very similar to your setup.

There are 2 subnets (1 and 2 see the IPs above) defined in "dhcpd.conf" (subnet 2 definition above, subnet 1 is basically the same, only printer is given fixed IP). "pf.conf" has the same setup for both subnets which is why I find this behavior strange.

[J65nko]

The ure1 ethernet adapter is an USB device. I wonder whether that could be the cause.

[erpete]

Originally Posted by J65nko

The ure1 ethernet adapter is an USB device. I wonder whether that could be the cause.

Yes, it is -- out of necessity. I'll try to swap ure1 and re0 and see if that helps. Thanks for the idea.

[erpete]

Connecting wifi router to internal NIC didn't help either.

Finally, upgrade to 7.0 seems to fix it.