Problems with my ASUS RT-AC87U Router when using another as an AP

[Demonos]

For the past 8 months I've been chasing different issues on and off. Almost every time I have an issue, flashing the latest Merlin firmware and resetting to factory defaults solves the issue for many months. That is, however, till this recent incident.

I own two ASUS RT-AC87U units. One is my core router in the basement, attached to my ARRIS / Motorola SurfBoard SB6121 DOCSIS 3.0 Cable Modem, as well as an unmanaged 24 port switch. The second unit is in Access Point mode and is sitting in my living room, which has a few gaming consoles wired to it as well as a Roku 3.

Both of these units have recently been flashed with Asuswrt-Merlin 378.51 (more than once) and reset to factory defaults each time. If I leave the core router running by itself, it can stay running indefinitely without any issues.

However when I turn on the AP, things get screwy. At first there are no issues. I can use either unit's WiFi or wired connections to access my LAN and the internet. For reasons beyond my understanding though, when I wake up at 6am every day, my router ceases to do any routing whatsoever. I can connect to the WiFi channels, but I cannot access my LAN, the Internet, or either unit. If I shut down the AP in the living room, my connection immediately regenerates and works again.

Lately I've been leaving the AP off until I can determine the cause. I pulled the logs from both units. First I looked at the AP which I assume is the cause of my problems.

Mar 22 11:21:41 rc_service: ntp 543:notify_rc restart_diskmon
Mar 22 10:21:50 RT-AC87U: start httpd
Mar 22 11:21:50 crond[578]: crond: crond (busybox 1.20.2) started, log level 8
Mar 22 11:21:51 disk monitor: be idle
Mar 22 11:21:51 kernel: gro enabled with interval 2
Mar 22 11:21:51 syslog: Generating SSL certificate...
Mar 22 10:21:57 Samba Server: daemon is started
Mar 22 11:22:30 qtn: bootcfg.tgz exists
Mar 22 11:22:31 disk monitor: be idle
Mar 22 15:17:55 kernel: br0: received packet on eth1 with own address as source address
Mar 22 20:37:34 kernel: br0: received packet on eth1 with own address as source address

The last two lines drew my attention. Received packet on eth1 with own address as source address. I have no idea why that would happen. So next I went to the core router's logs.

Mar 23 06:42:04 kernel: net_ratelimit: 585434 callbacks suppressed
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:43:14 dnsmasq-dhcp[913]: DHCPREQUEST(br0) 192.168.1.197 20:cf:30:51:a2:e8
Mar 23 06:43:14 dnsmasq-dhcp[913]: DHCPACK(br0) 192.168.1.197 20:cf:30:51:a2:e8 Teletron1

The log only goes back for 20 minutes because the router was flooded with these packets so much. The Router is set to 192.168.1.1 . DHCP starts at 1.4. the AP is set as 192.168.1.2. So I'm not overlapping IPs or anything. I trained as a Network/Security Engineer, but this is starting to frustrate me. I've spent a lot of time trying to get that AP to be happy and am rather tired of "working" when I'm at hope because these devices aren't playing nicely with each other. Anyone have any ideas? These units are brand new, since I moved into our new home back in the fall of 2014. Hardware designed for home users like this should really just work without a lot of issues.

 

[BartPl]

did you try two submasks?
router one 192.168.1.x
router two 192,168,2.x

 

[netwrks]

For the past 8 months I've been chasing different issues on and off. Almost every time I have an issue, flashing the latest Merlin firmware and resetting to factory defaults solves the issue for many months. That is, however, till this recent incident.

I own two ASUS RT-AC87U units. One is my core router in the basement, attached to my ARRIS / Motorola SurfBoard SB6121 DOCSIS 3.0 Cable Modem, as well as an unmanaged 24 port switch. The second unit is in Access Point mode and is sitting in my living room, which has a few gaming consoles wired to it as well as a Roku 3.

Both of these units have recently been flashed with Asuswrt-Merlin 378.51 (more than once) and reset to factory defaults each time. If I leave the core router running by itself, it can stay running indefinitely without any issues.

However when I turn on the AP, things get screwy. At first there are no issues. I can use either unit's WiFi or wired connections to access my LAN and the internet. For reasons beyond my understanding though, when I wake up at 6am every day, my router ceases to do any routing whatsoever. I can connect to the WiFi channels, but I cannot access my LAN, the Internet, or either unit. If I shut down the AP in the living room, my connection immediately regenerates and works again.

Lately I've been leaving the AP off until I can determine the cause. I pulled the logs from both units. First I looked at the AP which I assume is the cause of my problems.

Mar 22 11:21:41 rc_service: ntp 543:notify_rc restart_diskmon
Mar 22 10:21:50 RT-AC87U: start httpd
Mar 22 11:21:50 crond[578]: crond: crond (busybox 1.20.2) started, log level 8
Mar 22 11:21:51 disk monitor: be idle
Mar 22 11:21:51 kernel: gro enabled with interval 2
Mar 22 11:21:51 syslog: Generating SSL certificate...
Mar 22 10:21:57 Samba Server: daemon is started
Mar 22 11:22:30 qtn: bootcfg.tgz exists
Mar 22 11:22:31 disk monitor: be idle
Mar 22 15:17:55 kernel: br0: received packet on eth1 with own address as source address
Mar 22 20:37:34 kernel: br0: received packet on eth1 with own address as source address

The last two lines drew my attention. Received packet on eth1 with own address as source address. I have no idea why that would happen. So next I went to the core router's logs.

Mar 23 06:42:04 kernel: net_ratelimit: 585434 callbacks suppressed
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:42:04 kernel: br0: received packet on vlan1 with own address as source address
Mar 23 06:43:14 dnsmasq-dhcp[913]: DHCPREQUEST(br0) 192.168.1.197 20:cf:30:51:a2:e8
Mar 23 06:43:14 dnsmasq-dhcp[913]: DHCPACK(br0) 192.168.1.197 20:cf:30:51:a2:e8 Teletron1

The log only goes back for 20 minutes because the router was flooded with these packets so much. The Router is set to 192.168.1.1 . DHCP starts at 1.4. the AP is set as 192.168.1.2. So I'm not overlapping IPs or anything. I trained as a Network/Security Engineer, but this is starting to frustrate me. I've spent a lot of time trying to get that AP to be happy and am rather tired of "working" when I'm at hope because these devices aren't playing nicely with each other. Anyone have any ideas? These units are brand new, since I moved into our new home back in the fall of 2014. Hardware designed for home users like this should really just work without a lot of issues.

What happens when you disconnect the AP, and ping the AP's address?

 

[Demonos]

What happens when you disconnect the AP, and ping the AP's address?

Destination Host Unreachable.

I haven't tried setting the AP to 192.168.2.1 yet but I'm working on that now. I'll update my results in the morning. Really appreciate the extra set of eyes on this issue.

 

[Jong]

I have 1 AC87 router and 1 AP too and was experiencing exactly the same problem.

The problem you mention was fixed with the Official 4608 firmware that came out a couple of weeks ago. But yesterday 4950 was officially released and it is fantastic. I strongly suggest you update both the AP and Router.

I haven't played with Merlin's firmware, but if it is currently affected I am sure .52 will fix it soon.

 

[Demonos]

The problem still persists this morning, after setting the AP to 192.168.2.1. I'd post the logs but it looks the same as above, a flood of packets hitting my core router, making it impossible for other traffic to get anywhere.

The problem you mention was fixed with the Official 4608 firmware that came out a couple of weeks ago. But yesterday 4950 was officially released and it is fantastic. I strongly suggest you update both the AP and Router.

Bug fixes
- Fixed AP mode connection issue.

I'm guessing this is the fix you are referring to. I'll apply the latest release of 4950 tonight, until Merlin's next release with these fixes. I'll post another update tomorrow morning to at least confirm this was resolved. Thanks for the tip, Jong!

 

[Jong]

You don't need to do anything clever to set up your AP with 4950. Just regular AP Mode, with a static IP in the same subnet as the router is fine. It's probably best to assign control channels so the two do not interfere with each other.
Also, unless you have good reasons to do otherwise, use the same SSIDs for both AP and router, 2.4Ghz and 5Ghz. Devices roam nicely between all these networks as needed.

 

[Demonos]

You don't need to do anything clever to set up your AP with 4950. Just regular AP Mode, with a static IP in the same subnet as the router is fine. It's probably best to assign control channels so the two do not interfere with each other.
Also, unless you have good reasons to do otherwise, use the same SSIDs for both AP and router, 2.4Ghz and 5Ghz. Devices roam nicely between all these networks as needed.

I typically do all of this except for the SSIDs. No matter what I do, I see an SSID in my list for each router, even when the name is identical. Ideally I'd like to just have the one SSID for my whole network. However if they're both going to broadcast a separate SSID, I've been naming them differently so I at least know which one I'm connected to.

 

[Jong]

Very weird! What device? I've never seen that. Used the same SSID on several different APs/routers even from different manufacturers and certainly with my 2 AC87s the devices see only 1 SSID for all 4 radios - 2 on AP, 2 on Router. These devices range from the latest AC-mode laptops and iPhones to ancient 802.11b devices and much in between!

 

[Jong]

Edit: of course the passkey needs to be the same for all too.

 

[Demonos]

Very weird! What device?

I typically test the WiFi with my Nexus 5 or Nvidia Shield Tablet. It could have been a bug in an older firmware, or I could have set it up wrong. I'm totally willing to admit that.

Right now I have the AP set back to 192.168.1.2 and the Core Router set to 192.168.1.1, mostly so I can access both routers without having to force an IP in the .2 network. Both are running stock 4950. As per the instructions, I have reset both routers and reconfigured all my settings on both. Mostly DHCP, static IPs, WiFi channels, etc. Out of two routers and four channels, I will confirm I'm only seeing the one SSID now. Beamforming is working pretty well too. I get a smooth transition from one router to the other right around the center of the house, give or take 20 feet or standing around for 30 seconds, whichever come first.

Like last time, I'll let ya'll know how it goes in the morning when this issue typically occurs. Once again, thanks for the assistance!

 

[Jong]

Hope it all works out. It should. Personally I've turned off 5Ghz beamforming - my iDevices were disconnecting when sleeping and I think it was causing erratic performance with them too, but see how it works out for you.

Roaming Assist can help to nudge devices towards a better signal. I use it on my AP. I can't use it on my main router though because I have a guest network on there and I need to allow max range.

(NB: guest network on the APs DOES NOT WORK! It is like an alternate SSID for the main network - "guests" can access the main network and all packets on the network are rebroadcast on the alternate SSID drastically impacting wireless performance)

 

[Demonos]

After two days it seems the problem has been solved. No network degradation in the morning and the one SSID is working great. Thanks a lot for the help!

Good tip about the guest network, Jong. I usually enable it when we throw a party and I'd prefer to enable it on the AP since it's where guests usually congregate. I haven't needed it since the new year though so in the future I'll probably stick to enabling it on my core router. I'm probably going to camp out with this 4950 build for a while since it seems rather stable.

 

[Jong]

No problem. Glad it's all working out!

FYI, I have changed my mind on Roaming Assist. It works as advertised, but in my experience devices that need to maintain a connection e.g. playing media across the network, like iDevices and AirPlay, don't seem to best like being booted off a radio. While they will roam on their own and pretty much keep connections alive they can drop it when kicked off.