Multi-Router Mis-Configuration

[nietszch3]

Hi ispyisail
I have upgraded to 1.9.x. It worked right after the upgrade meaning routing for devices connected via WA854RE was correct. Then I configured DHCP leases. It still worked. Then I checked the box "FORCE DHCP leasing" in DHCP->static Ip leases. Routing was broken for devices connected via WA854RE.

I had to uncheck that box to get routing working again.

Perhaps this is something expected?

Eryk

[ispyisail]

now we are getting somewhere

you mean this tick box?

[nietszch3]

Hi

Yer, this is this setting.

[ispyisail]

Hi

Yer, this is this setting.

4 meg router? If so stick with 1.6.2

4 meg routers are end of life.

To make the firmware fit they had to take stuff out and that broke stuff.

[slow_speed]

@ nietszch3,
Look, don't worry about the extras. If it is working, then great! That's all you needed to do.

You did well. However, please post the exact steps you used to solve the problem, so the next guy knows.

[tehfink]

Hi ispyisail
I have upgraded to 1.9.x. It worked right after the upgrade meaning routing for devices connected via WA854RE was correct. Then I configured DHCP leases. It still worked. Then I checked the box "FORCE DHCP leasing" in DHCP->static Ip leases. Routing was broken for devices connected via WA854RE.

I had to uncheck that box to get routing working again.

I've also encountered this behavior: clients of an AP in client bridge mode lose ping & internet access once static DHCP addresses are enabled on the main router (even to clients which are statically assigned).

The router is a TP-Link TL-WR1043N/ND v2, the wireless client AP is a GL.iNet AR150, both running 1.9.X (Built 20170126-0747 git@ff413a5).

[Lantis]

Have you power cycled all devices?
The way the force dhcp check works is it makes sure that data coming from an IP address also comes from the Mac it is expecting.
This is verified by current leases vs assigned dhcp.

If the devices received an ip, and then you statically assigned another and enforced this rule, it would be locked out until it received a new and correct address

[tehfink]

Have you power cycled all devices?
The way the force dhcp check works is it makes sure that data coming from an IP address also comes from the Mac it is expecting.
This is verified by current leases vs assigned dhcp.

If the devices received an ip, and then you statically assigned another and enforced this rule, it would be locked out until it received a new and correct address

Thanks for the reply. Yes, I have tried power cycling all 3 devices, starting with the router, then the AP, then the client. No change, unfortunately.

The issue occurs even when the client does not use DHCP, but rather is setup with a static IP (the same IP as what should be given by the router's DHCP server).