2.4 GHz wireless disconnects, association issues

[acominym]

Ever since I upgraded to SDK6-based firmware, I have noticed that certain 2.4 GHz devices in my house will occasionally disconnect and be unable to reconnect until either they are rebooted (or their network interfaces are reset) or the router is rebooted. This typically happens only after the router has been running for at least a day or two.

I'm not sure if this is related, but generally, after a day or two of uptime, I'm also seeing the following message pop up in the router's log every couple of hours:

kernel: eth1: received packet with  own address as source address

This occurs on both ASUS' stock firmware and Merlin's custom builds.

Does anyone have any idea what might be going on?

EDIT: Forgot to mention: this is on an RT-N66U.

 

[jlake]

Try the troubleshooting steps in link below. If you already have the latest firmware, you can start with step 2.

http://forums.smallnetbuilder.com/showthread.php?t=12453

 

[acominym]

Hmm, I had already tried all of those steps at various points except for changing the channel width to 20 MHz. I'll try that and see if it works, though it may be a day or two before I can tell for sure if the problem is gone.

Out of curiosity, why does the channel width need to be forced down?

 

[RogerSC]

Hmm, I had already tried all of those steps at various points except for changing the channel width to 20 MHz. I'll try that and see if it works, though it may be a day or two before I can tell for sure if the problem is gone.

Out of curiosity, why does the channel width need to be forced down?

Because the router may try the 40MHz channel width, then fall back to 20MHz if that doesn't work, which can happen repeatedly. So what you see is unstable behavior. Very few areas have so few neighboring wireless networks that they have a couple of clear, non-overlapping channels to get the 40MHz channel width that you need. So if you just set the channel width to 20MHz, you'll see much more stable, consistent wireless behavior.

The other thing is that since you have a RT-N66U, you should try the SDK5 version of RMerlin's firmware. It's in the same place, in the "Experimental" folder. The SDK6 firmware works better with the RT-AC66U, as it turns out. I hear that Asus is working on making the SDK6 firmware work better with the RT-N66U, but until then the SDK5 firmware will probably work better for you.

 

[acominym]

Because the router may try the 40MHz channel width, then fall back to 20MHz if that doesn't work, which can happen repeatedly. So what you see is unstable behavior. Very few areas have so few neighboring wireless networks that they have a couple of clear, non-overlapping channels to get the 40MHz channel width that you need. So if you just set the channel width to 20MHz, you'll see much more stable, consistent wireless behavior.

Ah, I see. I live in a remote-ish area, so there aren't any measurable WiFi signals around other than the ones generated by my router, so I'm not sure what would be causing interference, but I guess it doesn't hurt to be safe.

The other thing is that since you have a RT-N66U, you should try the SDK5 version of RMerlin's firmware. It's in the same place, in the "Experimental" folder. The SDK6 firmware works better with the RT-AC66U, as it turns out. I hear that Asus is working on making the SDK6 firmware work better with the RT-N66U, but until then the SDK5 firmware will probably work better for you.

I've seen that build, but I have been reluctant to flash it because I'm not keen on reconfiguring my router from scratch yet again (and again if ASUS releases an improved SDK6 driver in the near future). Also, I have to admit the "experimental" label made me a little nervous . If changing the channel width doesn't end up fixing these issues, I will probably try it, though.

Thanks everyone for the responses.

 

[marnold]

I noticed the same behavior when I upgraded to 3.0.0.4.374.720. Backing off to 3.0.0.4.276 fixed it.

 

[jim769]

I noticed the same behavior when I upgraded to 3.0.0.4.374.720. Backing off to 3.0.0.4.276 fixed it.

Yes those older versions work great but they don't have a ipv6 firewall and that's not good. In fact to this day Asus still has not implemented a firewall for ipv6 but Merlin has. I believe starting with the 3.0.0.4.372.32_beta1 firmware that's what I have been using after trying them all on my N66 this build is the most stable for me no reboots no speed or signal issues just a great router that routes.

 

[RMerlin]

Yes those older versions work great but they don't have a ipv6 firewall and that's not good. In fact to this day Asus still has not implemented a firewall for ipv6 but Merlin has. I believe starting with the 3.0.0.4.372.32_beta1 firmware that's what I have been using after trying them all on my N66 this build is the most stable for me no reboots no speed or signal issues just a great router that routes.

That's why there's an SDK5 release...

 

[Aseuess]

2.4Ghz Wireless disconnects unstable and reproducible

Ever since I upgraded to SDK6-based firmware, I have noticed that certain 2.4 GHz devices in my house will occasionally disconnect and be unable to reconnect until either they are rebooted (or their network interfaces are reset) or the router is rebooted. This typically happens only after the router has been running for at least a day or two.

I'm not sure if this is related, but generally, after a day or two of uptime, I'm also seeing the following message pop up in the router's log every couple of hours:

kernel: eth1: received packet with  own address as source address

This occurs on both ASUS' stock firmware and Merlin's custom builds.

Does anyone have any idea what might be going on?

EDIT: Forgot to mention: this is on an RT-N66U.

I have observed the exact same behavior only in Access Point mode on multiple AC66U units (so it's possibly not the hardware). It is reproducible on a number of stock and Merlin firmwares. I have a couple of additional facts for others to see if there is a commonality.

- Reproducible on two separate AC66U units tried so far,

- Occurs reproducibly after a few hours of working fine on Asus stock firmware revs: 3.0.0.4.374_130 , 3.0.0.4.374_726 , 3.0.0.4.374_979 and Merlin 3.0.0.4_374.33 and 3.0.0.4_374.34

- The 5GHz signal will eventually go unstable also given a day or two

- ASUS 2.4Ghz is on channel 11, 20Mhz only

- On an ATT UVERSE based network with a TWOWIRE 3800HGV-B gateway, the 2.4Ghz signal from this gateway is on channel 1, hardware version 2700-100531-006, Software Version 6.9.1.42-plus.tm

- Even if the AC66U goes unstable, it is accessible via LAN and shows the other devices on the LAN correctly through the AC66U GUI

- The WiFi signals are still being transmitted when the system goes unstable.

- Normal connection speed is a rock solid steady 7.5MB with a 48ms +/- 3ms ping to outside but when unstable the speed is highly variable aroud only 800Kb with ping times of several hundred ms if one can still make a connection.

The following message is observed in the log even before the instability occurs and so far always in conjunction with instability:

kernel: eth1: received packet with own address as source address

- Same or differing SSID's from base UVERSE router doesn't matter.

- Letting the unit automatically get its IP address or manually entering it as an Access Point does not make a difference. It will go unstable under both conditions.

- There is nothing plugged into the routers output ports. Plugging the LAN into the WAN socket or the output ports doesn't matter. It will go unstable either way.

- We have used other vendors WiFi routers in the same physical location as Access Points without issue so I doubt it is interference from neighboring WiFi signals.

- Once it goes unstable, rebooting everything doesn't necessarily fix the problem every time! (Still trying to understand if possibly the UVERSE Router is temporarily corrupted by this AC66U instability.)

- No problems observed in Router or Media Bridge mode

Is there anyone else with anything in common?

 

[acominym]

I have observed the exact same behavior only in Access Point mode on multiple AC66U units (so it's possibly not the hardware). It is reproducible on a number of stock and Merlin firmwares. I have a couple of additional facts for others to see if there is a commonality.

[...]

- ASUS 2.4Ghz is on channel 11, 20Mhz only

That's interesting, because forcing the channel width down to 20 MHz completely eliminated this issue for me, with no other modifications to the default wireless settings aside from SSID and encryption.

The following message is observed in the log even before the instability occurs and so far always in conjunction with instability:

kernel: eth1: received packet with own address as source address

These log messages also disappeared for me when I reduced my 2.4 GHz channel width, so it seems they are indeed somehow connected to the instability.

The odd thing is that eth1 is the 2.4 GHz interface, and yet you say your 5 GHz becomes unstable, too? Are these messages associated with instability on either frequency or just 2.4 GHz?

 

[Aseuess]

That's interesting, because forcing the channel width down to 20 MHz completely eliminated this issue for me, with no other modifications to the default wireless settings aside from SSID and encryption.



These log messages also disappeared for me when I reduced my 2.4 GHz channel width, so it seems they are indeed somehow connected to the instability.

The odd thing is that eth1 is the 2.4 GHz interface, and yet you say your 5 GHz becomes unstable, too? Are these messages associated with instability on either frequency or just 2.4 GHz?

The 2.4GHz signal is now unstable on our test AC66U just 3 hours after a reset and there are now scores of these eth1 errors in the system log. Will keep an eye on it and see what the logs display if/when the 5GHz signal goes unstable.

 

[Aseuess]

A few more data points to this issue.

The 5GHz channel also eventually becomes unstable after a while on multiple AC66U routers.

I have not observed any entries in the log referencing eth2 at any time.

I have also observed situations where the 2.4GHz and 5GHz channels spontaneously become fully operational again for a while. I note the following entry in the log around the time both channels were operational:
Nov 22 05:57:30 kernel: eth1: received packet with own address as source address
Nov 22 07:06:09 rc_service: udhcpc_lan 342:notify_rc stop_dnsmasq
Nov 22 07:06:09 rc_service: udhcpc_lan 342:notify_rc start_dnsmasq
Nov 22 09:29:45 kernel: eth1: received packet with own address as source address
Is the rc_service message coincidental or a clue? Unfortunately, both channels will again become unstable after a few hours.

 

[acominym]

A few more data points to this issue.

The 5GHz channel also eventually becomes unstable after a while on multiple AC66U routers.

I have not observed any entries in the log referencing eth2 at any time.

I have also observed situations where the 2.4GHz and 5GHz channels spontaneously become fully operational again for a while. I note the following entry in the log around the time both channels were operational:
Nov 22 05:57:30 kernel: eth1: received packet with own address as source address
Nov 22 07:06:09 rc_service: udhcpc_lan 342:notify_rc stop_dnsmasq
Nov 22 07:06:09 rc_service: udhcpc_lan 342:notify_rc start_dnsmasq
Nov 22 09:29:45 kernel: eth1: received packet with own address as source address
Is the rc_service message coincidental or a clue? Unfortunately, both channels will again become unstable after a few hours.

I suppose it makes sense that restarting dnsmasq would (temporarily) fix association issues, but I have no idea what would trigger the restart or what would cause the connection to fail again after a few hours. Someone more familiar with the internals of ASUS' firmware might be able to provide more information here.

 

[Aseuess]

Anyone using AC66U successfully in Access Point mode?

Another data point. The device in plain old router WiFi mode seems to still be working OK on my network after several days so it is unlikely that the 2wire Uverse base router or the ASUS RT-AC66U hardware is the issue in this case. It looks like Access Point mode is the problem. Although its pretty straightforward, is there anyone successfully using the AC66U in Access Point mode who could share what firmware and any configuration tips they used?

 

[Pericynthion]

Are you running separate SSIDs for 2.5 and 5ghz or are they the same name?


Sent from my iPhone using Tapatalk

 

[Aseuess]

Are you running separate SSIDs for 2.5 and 5ghz or are they the same name?

In Access Point mode, the system is unstable no matter what combination of the same or different SSID's between the 2.4GHz, 5GHz and base WiFi router are used. Shutting off the 5 GHz channel in Access Point mode will still result in the system eventually going unstable on 2.4GHz (and vis versa) as described above.

As a test, I am currently running the AC66U in WiFi router mode with the same SSID on it as the base router and it appears to work fine with mobile devices connecting and disconnecting between the two routers without issue (even though the AC66U WiFi subnet is the same range as the base router's WiFi and LAN DHCP. The mobile device gets a new IP in the range of whichever WiFi router's signal it is under. As expected due to the AC66U NAT involved in router mode, the LAN and WiFi connections controlled by the other router are not visible. A little not typical but working as expected in the router mode)