From: tnp@invalid.invalid
On 18/01/2026 10:54, Michael Schwingen wrote:
> On 2026-01-13, mm0fmf wrote:
>> I may be wrong and sometimes am but I thought all you needed to add was
>> edit /etc/sysctl.conf and ensure it contains "net.ipv4.ip_forward = 1",
>> save and reboot. Mine came with the line in there but commented out.
>
> That is for IP routing, using seperate IP networks on the ethernet and wifi
> side. You can do that, but it may make the setup more complicated and cause
> problems with prototols that rely on broadcast/multicast packets, as these
> are not routed. Also, if your WIFI clients require internet access, you need
> to setup routes on your internet router so it can forward packages back to
> the WIFI gateway. You need to decide if routing or bridging best suits your
> needs.
>
> "normal" WIFI access points usually operate in bridging mode.
>
>
>
> For bridging, you need to set up a bridge device, with both the ethernet and
> wifi devices slaved to the bridge. In that scenario, the slave devices do
> not get IP addresses assigned - the bridge device is the one with the IP
> address.
>
> Looking at a running example (on custom hardware, not a raspberry),
> with a single of the 3 wifi modules active, it looks like this (eth1 is the
> ethernet interface, phy0-ap0 is wifi):
>
> root@lx6500-dev:~# brctl show
> bridge name bridge id STP enabled interfaces
> br-lan 7fff.00a057802bee no eth1
> phy0-ap0
>
> root@lx6500-dev:~# ip l
> 4: eth1: mtu 1500 qdisc mq master br-lan
state UP mode DEFAULT group default qlen 1000
> link/ether 00:a0:57:80:2b:ee brd ff:ff:ff:ff:ff:ff
> 5: br-lan: mtu 1500 qdisc noqueue state UP
mode DEFAULT group default qlen 1000
> link/ether 00:a0:57:80:2b:ee brd ff:ff:ff:ff:ff:ff
> 6: phy0-ap0: mtu 1500 qdisc noqueue
master br-lan state DOWN mode DEFAULT group def0
> link/ether 04:f0:21:bf:45:cc brd ff:ff:ff:ff:ff:ff
>
> root@lx6500-dev:~# ip a
> 4: eth1: mtu 1500 qdisc mq master br-lan
state UP group default qlen 1000
> link/ether 00:a0:57:80:2b:ee brd ff:ff:ff:ff:ff:ff
> 5: br-lan: mtu 1500 qdisc noqueue state UP
group default qlen 1000
> link/ether 00:a0:57:80:2b:ee brd ff:ff:ff:ff:ff:ff
> inet 192.168.1.1/24 brd 192.168.1.255 scope global br-lan
> valid_lft forever preferred_lft forever
> inet6 fe80::2a0:57ff:fe80:2bee/64 scope link proto kernel_ll
> valid_lft forever preferred_lft forever
> 6: phy0-ap0: mtu 1500 qdisc noqueue
master br-lan state DOWN group default qlen 1000
> link/ether 04:f0:21:bf:45:cc brd ff:ff:ff:ff:ff:ff
>
> Using that configuration, WIFI clients get addresses from the existing DHCP
> server on the LAN, and are in the same IP network as the LAN devices.
>
> cu
> Michael
Yes. a bridge is - a bridge! and that is how normal wifi access points work
My issue is not with the basic config. It is with the performance - and
in fact a subset of the performance - access to my LAN is FINE access
the internet beyond is dire, but not impossible
For reference this is what IP link shows:
# ip l
1: lo: mtu 65536 qdisc noqueue state UNKNOWN mode
DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: mtu 1500 qdisc mq master
nm-bridge state UP mode DEFAULT group default qlen 1000
link/ether d8:3a:dd:85:22:b1 brd ff:ff:ff:ff:ff:ff
3: wlan0: mtu 1500 qdisc pfifo_fast
master nm-bridge state UP mode DORMANT group default qlen 1000
link/ether d8:3a:dd:85:22:b2 brd ff:ff:ff:ff:ff:ff
4: nm-bridge: mtu 9000 qdisc noqueue
state UP mode DEFAULT group default qlen 1000
link/ether d8:3a:dd:85:22:b1 brd ff:ff:ff:ff:ff:ff
Everything seems as it should be. No interface reports packet loss, yet
still it is massive but *only when going to/from wifi client to.from the
Internet*.
I do not understand what is so different about a packet going to or from
the internet that causes the problem.
--
“But what a weak barrier is truth when it stands in the way of an
hypothesis!”
Mary Wollstonecraft
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
