@AnnusKhan 

Hello again!

I just re-read the thread, and realized I'd made a typo where I'd suggested splitting the networks.  What I'd meant was the following:

Devices:  134.1.1.0/24

Clients:    134.1.4.0/18 (not 16) and I would suggest even making it 134.1.4.0/22 which still gives you a thousand IP's to hand out to your 750 users.

Next up, I have been working with support for months on what I perceive to be a memory leak issue in the software architecture used by the ER routers.  It is worst on the ER605v1 as it has by far the least amount of memory and so things tend to happen fastest and first on that model.  Essentially I have been seeing a linear increaes in real memory consumption over time, which eventually gets to the point where memory starvation occurs...when this happens, process crash...typically it's the Omada application which causes disconnects, but I believe I have seen DHCP die first as well.  So I"m recommending you reduce the pool of IP's for clients to reduce the amount of resources required on the router and controller to track them all.  I'm also going to suggest you perform a weekly reboot (the controller has the ability to configure the Reboot Schedule for devices, and reset just the ER7206 at 4am Saturday or whatever).  This and updating to the latest firmware should help eliminate the spinning 'Waiting for IP' stuff on clients.

Now to your new question, you wanted two policies...did you just mean you want to the two subnets to behave as expected, or do you have different gateways?  I need a little more clarity around exactly what you are asking for.  Also, a screen grab of your topology with Labels on would help me understand the hardware architecture, and a screen grab of all subnet definitions in the Wired Networks->LAN would be good too.

  @AnnusKhan Hi, we might need @Clive_A  or @Hank21 to wade in here, but the error sounds like somebody isn't connected to somebody else, whether that is your device is on a VLAN that doesn't have access to some component of the captive portal, or the portal itself is not getting a response from say a device gateway.  I don't use CP's so I"m a bit in the dark here.

Hi @AnnusKhan 

Thanks for posting in our business forum.

Is your original issue resolved or not? If no, I'll continue the original issue with you.

The new issue you said "unauth", is a controller issue which I don't know a lot about. I suggest you start a new thread in the Controller section.