Linksys WRT1900AC refuses to accept a different IP (192.168.0.2) for itself

[Ichinisan]

My question to anyone with this router:
WHY CAN'T WE GET THE ROUTER TO USE "192.168.1.2" AS ITS INTERNAL IP ADDRESS?

Spoke to a client today. He has a whole-home multi-room DVR system from Arris. The master "media gateway" device (Arris MG5225G) is basically an all-in-one:

• DOCSIS 3 cable modem
• NAT router
• WiFi access point
• MoCA bridge (coax networking)
• 5-tuner DVR with HDD storage
• MTA (VoIP phone)

Behind that, he has his own Linksys WRT1900AC router. This one:

Behind the Linksys device, he has 3 important systems:
- Gaming PC
- Server PC
- Network Drive (Western Digital)

He wants the server PC to be accessible to the Internet.

This is a scenario I've dealt with many times before. I would usually configure the router for AP mode. If it doesn't have the option, I would do it manually by disabling DHCP on the router and moving the router's incoming connection to one of its LAN ports. We don't have to do that because the router has a "bridge mode," which should basically make it behave as an access point instead of a NAT router. That's also good because all of his LAN ports are already occupied.

In his router, he had port forwarding configurations for America's Army, Battlefield 2, and FTP. We took a screenshot of these settings. Strangely, his router didn't seem to have DHCP reservations for those devices...but they had the correct addresses for his port forwarding rules.

The default settings for each device:

Arris device: 192.168.0.1
DHCP start: 192.168.0.2

Linksys device: 192.168.1.1
DHCP start: 192.168.1.100​

I knew that simply enabling bridge mode would probably prevent us from logging-in to the Linksys router settings. I wanted to make it so he could access the settings of both devices:

Arris device: 192.168.0.1
Linksys device: 192.168.0.2

We changed the Arris device to use 192.168.0.3 as the starting DHCP addres. Even though the DHCP client list didn't show anything using the .2 address, we rebooted it so all devices would pull a new IP address. In the Linksys, we set the internal IP to 192.168.0.2 and disabled DHCP. We applied the changes, and this is where things get strange...

The computer pulled a 10.231.180.x IP address
We copied down the default gateway (also 10.231.180.x) that was pulled via DHCP and used it to log-in to the Linksys router. Once we got in, the router was clearly showing DHCP enabled and the subnet was that strange IP range that was NOT the IP range I had configured.

To take a different approach, we went into the Arris device and changed it from 192.168.0.1 to 192.168.1.1, then set the DHCP range to start with 192.168.1.4 -- then saved and rebooted the device.

We attempted to configure the Linksys device as 192.168.1.2 -- to make sure we wouldn't lose the ability to log-in to the Linksys if it decided to use yet another random IP address, I deliberately left DHCP active this time (to be disabled later) with a starting IP of 192.168.1.100 -- which shouldn't conflict with any of the current DHCP leases on the Arris that all have low numbers.

Before applying these changes, while clicking back-and-forth between fields in the settings we saw the UI glitch by changing the router's IP from "192.168.1.2" to "192.168.180.2" -- We re-typed it 192.168.1.2 again and tried to duplicate the UI glitch, but it didn't happen any more. So we finally applied the change.

Then the computer pulled yet another strange IP address of 10.117.154.x and default gateway of 10.117.154.134 (!?!?). We logged-in to the router again with the default gateway address and it shows that strange brand new IP range -- NOT what I had configured. Another thing of note: The router's own IP lies within its DHCP range, meaning it could hand that address to a LAN device and cause a conflict with itself!

Ultimately, we configured bridge mode in the Linksys router. Because the router refuses to take the 192.168.1.2 address, we can no longer log-in and change WiFi settings (unless we do a factory reset). Hopefully we won't ever need to...but this bothers me and it bothers the client.

I configured the Arris device with DHCP reservations to give his 3 devices the same IPs they had before and did the same port forwarding configuration in the Arris device. It appears to be working.

 

[Fardringle]

Either the Linksys router is defective and randomly changing its settings (maybe try updating the firmware) or something else on the network is assigning those 10.x IP addresses to it.

 

[VirtualLarry]

You can't have the same subnet on the WAN port as the local LAN IP / LAN ports.

The router was detecting the 192.168.0.1 on the WAN port, and refusing to configure its LAN IP as 192.168.0.2. If it did, it wouldn't know how to router between WAN and LAN. It does this for your own protection.

If you want to daisy-chain routers, assign the primary router (connected to modem, or integrated with it), to 192.168.0.1, and then assign the secondary router (the Linksys)'s WAN as a static WAN IP, and make it 192.168.0.2. Then assign the secondary router's LAN IP to 192.168.1.1, and set up the DHCP server and allocation for that subnet for the LAN devices.

You can still access the LAN IP of the primary router this way, and access its config pages.

The only other change you need to make, if you use port forwarding or VPN server or any advanced features of the secondary router, then you need to add the secondary router's WAN IP into the DMZ setting of the primary router.

I use this very setup myself, so I'm very familiar with it.

My primary router is a FIOS ActionTec, and my secondary router is an Asus running Tomato.

 

[Fardringle]

Ah, good point. Reading most of the setup I just assumed this was being set up as an access point. To set it up in a double-NAT, they have to be different subnets as VirtualLarry said.

 

[VirtualLarry]

Ah, good point. Reading most of the setup I just assumed this was being set up as an access point. To set it up in a double-NAT, they have to be different subnets as VirtualLarry said.

My setup isn't double-NAT, per se. The DMZ feature makes the second NAT basically transparent, as far as port-forwarding and the like are concerned.

 

[Ichinisan]

Either the Linksys router is defective and randomly changing its settings (maybe try updating the firmware) or something else on the network is assigning those 10.x IP addresses to it.

It's definitely the Linksys router. The default gateway revealed by ipconfig leads us to the same setup page. He only has 1 Linksys router.

You can't have the same subnet on the WAN port as the local LAN IP / LAN ports.

The WAN interface shouldn't have an IP at all once bridge/AP mode is enabled.

The router was detecting the 192.168.0.1 on the WAN port, and refusing to configure its LAN IP as 192.168.0.2. If it did, it wouldn't know how to router between WAN and LAN. It does this for your own protection.

I figured as much. With any other router, I could set the internal IP and then enable bridge mode.

If you want to daisy-chain routers, assign the primary router (connected to modem, or integrated with it), to 192.168.0.1, and then assign the secondary router (the Linksys)'s WAN as a static WAN IP, and make it 192.168.0.2. Then assign the secondary router's LAN IP to 192.168.1.1, and set up the DHCP server and allocation for that subnet for the LAN devices.

I'm trying to disable the DHCP server in the secondary router and have it function as an AP bridge. That's done, but I can't do it in a way that leaves it with an accessible LAN IP address on the same subnet so settings can be configured.

You can still access the LAN IP of the primary router this way, and access its config pages.

Yeah. I told him he shouldn't need to access the Linksys settings unless he suddenly wants to change his WiFi name and password. He said he doesn't want to change that, but it bothers him that he can't get into the settings.

The only other change you need to make, if you use port forwarding or VPN server or any advanced features of the secondary router, then you need to add the secondary router's WAN IP into the DMZ setting of the primary router.

We were specifically doing AP/bridge mode to avoid double-NAT.

 

[Ichinisan]

My setup isn't double-NAT, per se. The DMZ feature makes the second NAT basically transparent, as far as port-forwarding and the like are concerned.

The disadvantage is that devices on one network might not see devices on the other. For instance, the Arris DVR remote control app that this client might want to use...

If we configure the Linksys to behave as an AP/bridge, it wouldn't matter which device he connects through. It's one big LAN and all devices can see each other.

 

[VirtualLarry]

The disadvantage is that devices on one network might not see devices on the other. For instance, the Arris DVR remote control app that this client might want to use...

Yes, devices connected to my FIOS router cannot access my LAN. This is by design.

If we configure the Linksys to behave as an AP/bridge, it wouldn't matter which device he connects through. It's one big LAN and all devices can see each other.

Why do you keep referring to it as "AP/Bridge". Those two roles are very different.
AP = use LAN port, connect via wifi.
Bridge = use WAN and LAN ports, if NATted bridge, then NAT between WAN and LAN.