2.4G WIFI connectivity problem

[niXta]

Hi,

I have a Chinese AC86U with open region settings set to Europe running merlin 384.5. All good so far.

I've been having some problems with my 2.4G network disappearing/unable to connect.
On top of that, 2 or 3 times, the webserver freezes (Always when the 2.4G is down). When it does, SSH is still accessible.

I have some logs I would like to share, I have masked my MAC's, is there anything else I should mask before posting? SN?

Log deleted

 

[ColinTaylor]

The only thing you need to mask is your WAN IP address and any user names and passwords. You don't need to mask MAC addresses.

EDIT: You said in your previous post that you hade the same problem with a different router and it was caused by one of your clients.

 

[niXta]

The only thing you need to mask is your WAN IP address and any user names and passwords. You don't need to mask MAC addresses.

EDIT: You said in your previous post that you hade the same problem with a different router and it was caused by one of your clients.

Ok!
Will search and put it up.
I thought I fixed it by restoring my Shield TV, unfortunately it is still acting up - but not in the exact same way. It never says wrong password now. Some devices can’t connect and the GUI is unreachable sometimes.


Skickat från min iPhone med Tapatalk

 

[niXta]

The only thing you need to mask is your WAN IP address and any user names and passwords. You don't need to mask MAC addresses.

EDIT: You said in your previous post that you hade the same problem with a different router and it was caused by one of your clients.

Ok, here is the log. I couldn't connect devices on the 2.4G and couldn't access the GUI.
So I ssh'ed to it and rebooted and pulled the log.

 

[ColinTaylor]

I don't know the solution but your problem seems to be that the router's Mastiff process is constantly crashing and being restarted every 60 seconds .

May 16 11:58:25 rc_service: watchdog 808:notify_rc stop_aae
May 16 11:58:25 rc_service: watchdog 808:notify_rc start_mastiff
May 16 11:58:25 rc_service: waitting "stop_aae" via watchdog ...
May 16 11:58:25 NAT_Tunnel: AAE Service is stopped
May 16 11:58:25 NAT_Tunnel: AAE Service is stopped
May 16 11:58:26 NAT_Tunnel: AAE Service is stopped
May 16 11:58:26 AAE: AAE Service is started
May 16 11:58:26 Mastiff: init
May 16 11:59:03 kernel: mastiff[23453]: unhandled level 2 translation fault (11) at 0x36223a2e, esr 0x92000006
May 16 11:59:03 kernel: pgd = ffffffc014e4e000
May 16 11:59:03 kernel: [36223a2e] *pgd=0000000017258003, *pud=0000000017258003, *pmd=0000000000000000
May 16 11:59:03 kernel: CPU: 0 PID: 23453 Comm: mastiff Tainted: P           O    4.1.27 #2
May 16 11:59:03 kernel: Hardware name: Broadcom-v8A (DT)
May 16 11:59:03 kernel: task: ffffffc0193b2b40 ti: ffffffc01578c000 task.ti: ffffffc01578c000
May 16 11:59:03 kernel: PC is at 0xf6c99218
May 16 11:59:03 kernel: LR is at 0x0
May 16 11:59:03 kernel: pc : [<00000000f6c99218>] lr : [<0000000000000000>] pstate: 60010010
May 16 11:59:03 kernel: sp : 00000000ffa27e20
May 16 11:59:03 kernel: x12: 00000000ffa27e44
May 16 11:59:03 kernel: x11: 00000000f6d667a8 x10: 00000000f6d667d8
May 16 11:59:03 kernel: x9 : 0000000036223a22 x8 : 0000000000000010
May 16 11:59:03 kernel: x7 : 00000000008bea88 x6 : 00000000008bea78
May 16 11:59:03 kernel: x5 : 0000000000000070 x4 : 00000000008bea08
May 16 11:59:03 kernel: x3 : 00000000f6007d22 x2 : 00000000000000e9
May 16 11:59:03 kernel: x1 : 0000000000000000 x0 : 0000000000000000
May 16 11:59:26 rc_service: watchdog 808:notify_rc stop_aae
May 16 11:59:26 rc_service: watchdog 808:notify_rc start_mastiff
May 16 11:59:26 rc_service: waitting "stop_aae" via watchdog ...

The last reboot also had this crash dump, but I suspect it's coincidental.

Feb 14 01:00:17 kernel: jffs2: notice: (1) check_node_data: wrong data CRC in data node at 0x001e9f9c: read 0x4b16d2df, calculated 0x61f866ff.
Feb 14 01:00:17 kernel: _ Reboot message ... _______________________________________________________
Feb 14 01:00:17 kernel: Emergency Sync completebr0: port 6(eth6) entered disabled stateBUG: recent printk recursion!br0: port 5(eth5) entered disabled statebr0: port 4(eth4) entered disabled statebr0: port 3(eth3) entered disabled statedevice eth1 left promiscuous modebr0: port 1(eth1) entered disabled statedevice eth2 left promiscuous modebr0: port 2(eth2) entered disabled statedevice eth3 left promiscuous modebr0: port 3(eth3) entered disabled statedevice eth4 left promiscuous modebr0: port 4(eth4) e
Feb 14 01:00:18 kernel: , max. volumes count: 128ubi1: max/mean erase counter: 3/2, WL threshold: 4096, image sequence number: 1960956644ubi1: available PEBs: 0, total reserved PEBs: 64, PEBs reserved for bad PEB handling: 4ubi1: background thread "ubi_bgt1d" started, PID 18743UBIFS (ubi1:0): UBIFS: mounted UBI device 1, volume 0, name "nvram", R/O modeUBIFS (ubi1:0): LEB size: 126976 bytes (124 KiB), min./max. I/O unit sizes: 2048 bytes/2048 bytesUBIFS (ubi1:0): FS size: 5840896 bytes (5 MiB, 46 LEBs)
Feb 14 01:00:18 kernel: core: deregistering interface driver rndis_hostusbcore: deregistering interface driver cdc_etherusbcore: deregistering interface driver ax88179_178ausbcore: deregistering interface driver asixusbcore: deregistering interface driver cdc_acmcdrom: Uniform CD-ROM driver unloadedusbcore: deregistering interface driver usblpdhd_detach(): thread:dhd_watchdog_thread:23e terminated OKFPM Pool 1: invalid token 0x226e8000 freedFPM Pool 1: invalid token 0x22250000 freedFPM Pool 1: invalid
Feb 14 01:00:18 kernel: ool 1: invalid token 0x20480000 freed
Feb 14 01:00:18 kernel: ____________________________________________________________________________

I don't know what Mastiff is but it seems to be linked to "NAT_Tunnel". Do you have any options on the router to turn these things off?

 

[niXta]

Mastif? They are cute dogs [emoji23] /jk
I only know of the Asus NAT, should I turn it off?


Skickat från min iPhone med Tapatalk

 

[ColinTaylor]

I only know of the Asus NAT, should I turn it off?

I don't know. NAT is not the same as NAT_Tunnel (AFAIK).

The same error was reported here by @Blacklistedcard. What firmware version are you running?

https://www.snbforums.com/threads/p...5-early-test-builds.45769/page-18#post-401394

 

[niXta]

I don't know. NAT is not the same as NAT_Tunnel (AFAIK).

The same error was reported here by @Blacklistedcard. What firmware version are you running?

https://www.snbforums.com/threads/p...5-early-test-builds.45769/page-18#post-401394

I'm using merlin 384.5. Hmm, debugging this doesn't look promising...

Any advice where to go from here? Reset and set up manually for the 32 time?

 

[niXta]

I don't know. NAT is not the same as NAT_Tunnel (AFAIK).

The same error was reported here by @Blacklistedcard. What firmware version are you running?

https://www.snbforums.com/threads/p...5-early-test-builds.45769/page-18#post-401394

I turned off "Asusnat tunnel" and haven't seen the message since.

I'm not sure if my 2.4G/GUI problems are fixed too but I'll soon find out...

 

[Treadler]

I turned off "Asusnat tunnel" and havn't seen the message since.

I'm not sure if my 2.4G/GUI problems are fixed too but I'll soon find out...

Asusnat tunnel > disable.
Merlin advised this some time ago, I think he said it used CPU & ram, & wasn’t sure what it did.
I have had it turned off since the option was there to do so.

 

[Grisu]

And why do we have this problem only in 2,4G?
How to turn it off with asuswrt in AP-mode?

 

[agilani]

In my case i am still not sure why it only happens on 2.4ghz. It looks like maybe some sort of dhcp broadcast storm. Where certain devices ask for a dhcp address over and over. Literally, my debug syslog were full of dhcprequest and dhcpack over and over for the same devices. I also increased the dhcp time to two days to see if that lessens the symptom.

I tracked down the devices that were doing that in my logs and moved them to other ssid's and now my 2.4ghz has been rock solid. In one instance even changing the nic card from the device did not make any difference. I should have captured the network traffic to analyze it more, but now that my 2.4ghz is stable....i doubt I'll get to it.

 

[Grisu]

no solution for me, there is no dhcp-storm to see in the logs.

 

[agilani]

no solution for me, there is no dhcp-storm to see in the logs.

Did you turn on debug logging? Do you have the check box next to Hide DHCP/RA queries under dhcp server set to no?

 

[Grisu]

yes had it on debug-level but now back to stock asuswrt testing mesh, had been in AP-mode with DHCP on modem-router.

 

[niXta]

I've once again got major 2.4G interference. It seems to be coming from the Shield TV I have, it has been completely wiped and restored without any backups. Might be faulty hw?

I could in real time turn off and on and see all my 2.4G devices connect and disconnect.

 

[Grisu]

in my case there are no other devices but my 2 asus routers sending WiFi.

 

[niXta]

in my case there are no other devices but my 2 asus routers sending WiFi.

Ok, just to be clear. I don't believe I have the 2.4G WiFi problems as the rest in this thread anymore.
I had the same problems I have now with my R7000.

 

[Grisu]

seems to be true for you
I can even turn off 86U and have slow 2,4G on 68U or vice versa, but there are no other networks seen around me, whats going on here?
-25dbm from one and other -45dbm should give good speed over 100mbps, not <30 going down to 1 mbps fluctuating/hopping up after some second and slowing down again.

I had times where 68U had >130mbps with previous firmwares. 86U had all the time problems, so for me firmware disaster.

 

[niXta]

seems to be true for you
I can even turn off 86U and have slow 2,4G on 68U or vice versa, but there are no other networks seen around me, whats going on here?
-25dbm from one and other -45dbm should give good speed over 100mbps, not <30 going down to 1 mbps fluctuating/hopping up after some second and slowing down again.

I had times where 68U had >130mbps with previous firmwares. 86U had all the time problems, so for me firmware disaster.

Yep, it might be firmware issues.
I hate that A̶̶̶s̶̶̶u̶̶̶s̶̶̶ ̶B̶r̶o̶d̶c̶o̶m̶ ̶l̶o̶c̶k̶e̶d̶ ̶t̶h̶e̶ ̶b̶o̶o̶t̶l̶o̶a̶d̶e̶r̶ ̶s̶o̶ ̶y̶o̶u̶ ̶c̶a̶n̶'̶t̶ ̶r̶u̶n̶ ̶o̶t̶h̶e̶r̶ ̶f̶i̶r̶m̶w̶a̶r̶e̶'s̶ ̶o̶n̶ ̶t̶h̶e̶ ̶8̶6̶U̶...

Edit: Actually it's 3rd party firmware makers that can't/won't create firmware's for this platform.