[MB55R32]

Having issues with Random Reboots, I little info I've been using a Netgear R7800 for a few years

till recently about 3 months ago when the r7800 stopped powering on I tested with other PSU’s and same I had a backup Linksys MR8300,

I saw the thread on the Dynalink WRX36 AX Router and saw it was selling for a good price I got one and this is were my issues started,

I copied most of the settings from my MR8300 like I did with my R7800 and no I took screenshots of the setup and used those to setup the WRX36 Router,

For some reason the router will work fine for a few days and randomly reboot at bad times when I was working remotely,

I did troubleshooting re-did settings and updated to newer Firmware thanks to B.S and DDWRT team.

The further I got I noticed a connection with NSS and ECM features with QOS enabled with these newer AX Routers.

The most stable I can get is to have NSS and ECM features under - Setup Tab > Optional Settings
Enabled and SFE Disabled with QOS enabled it lasts around 2 to 3 days before it reboots randomly.

Testing to make the error occur more frequently use this settings

under Setup Tab > Optional Settings > Shortcut Forwarding Engine = Disable all features here,

QCA NSS Subsystem = Disable.

Under NAT / QOS Tab > QOS Settings >
Start QOS = Enabled
Port = WAN
Packet Scheduler = HTB
Queuing Discipline = FQ_Codel – same with CAKE
Download Link = 20000 Kbit/s
Uplink = 9000 Kbit/s

I use MAC Priority – I added a few MAC address and add
Download Link = 200000 Kbit/s
Uplink = 25000 Kbit/s

I have dozen devices so I use QOS throttle data on IOT and IP Phones that do no need the bandwidth

I can take screenshots my setup if needed, I even bought a new router thinking it was defective my return windows was over on the WRX36, On Ebay I saw good deal on Linksys MX4300 $38 shipped.

I thought this would be the cure but like the other router it does the same reboots randomly

I went back to using the MR8300 I think there is issues with AX routers with NSS system and QOS.

I attached the Dmesg from Admin tab ran the command on GUI and had this output

Sorry for the long post.

Forgot to add the important info


MX4300
Router Model
Linksys MX4300
Firmware Version
DD-WRT v3.0-r59582 std (02/09/25)
Kernel Version
Linux 6.6.75-rt29 #2465 SMP Sun Feb 9 05:37:04 +07 2025 aarch64

[lexridge]

While I have no idea what may be going on with your setup, I do know that you cannot use any form of SFE with QoS (this includes NSS-ECM too, afaik). I see you did test this with it turned off.

I have two MX4300s with up to 12 day uptimes but I am not using the WAN ports, so not a fair comparison. I am sure they could have had much longer uptimes, but that gets reset whenever I update to newer firmware ofc.

You should enable syslogd in Services and try to capture a log when the crash occurs. You will need to ssh or telnet into the router and run:
tail -f /var/log/messages
And hopefully you can capture some useful information.

[MB55R32]

Interesting update I did use tail -f /var/log/messages in ssh session,

I noticed the router reboots at a faster rate if I’m loading website with constant data, Reddit scrolling down for a few min’s and loading my browser with live feeds,

The router will crash I caught one around 26mins mark, I was disconnected and was auto switched to my MR8300 router,

The open SSH session did not pickup the reboot, I think I need to use a USB to TTY Serial Adapter to see live feed,

Also not sure if related but I did notice that under DD-WRT Settings: Status Tab > Router > Memory - Available / Total > Buffers - on my MR8300 its using around 8%

And moves down and up getting updated every few seconds, and under the MX4300 its at 0% I wonder if this might be an issue were I think the router runs out space for constant updates ?

On the screenshot MX4300 is on top and the MR8300 on the bottom on the Router status Tab.

[kernel-panic69]

I thought the defaults for conntrack max IP connections were changed on these. 4096 is a relatively LOW number.

[Per Yngve Berg]

It's 32768 on my R7800.

It's shown at the bottom of Status->Router

[mrjcd]

EA8500 running r59661

[MB55R32]

Ok, I was able to capture data of the USB Serial TTY adapter,

I got three files as the MX4300 rebooted with me not realizing, I used Putty and had it save a log file,

The first file I don't think got much but the 2nd and 3rd files I believe have some useful data, this is the one I noticed i was kicked off WiFi and the top LED light on router went from Solid Blue to Solid Red,

And the Router refused to work no WiFi connection it was like this for 10mins or so before I unplugged the power.

[Alozaros]

I know it sound like a cliche ...but..check your Power Adapter would be the first thing to check...

Otherwise..my WRX36 is fine with NSS-ECM only...(i don't use QoS)..but even thou...WRX36 has enough power to sustain a good WAN performance even without acceleration...

p.s. Also ive read some Amazon reviews of people complaining on a bad radio chip....so unless its not hardware issues..logs will be very valuable if its a software one...

[MB55R32]

Update to the random reboots, 2 weeks ago I had the MX4300 behind my MR8300 as double NAT and only had one device assigned,

with the same settings with a tweak on the TX power from 20 to 27dB,
I tested above settings and noticed no random reboots and under load it was 5 days without a reboot,

Thinking I had this resolved I tested the same settings on my WRX36 I made this the main Router,

Testing failed within 30 mins. into the testing, I than searched the forums for similar symptoms and found posts of discussing ACK Timing for disconnects and suggested shorter value or 0 for some setups of certain AP location,

I checked under Wireless Settings > Advanced settings > Sensitivity Range / ACK Timing

The default setting is 500 meters looking a discussions they suggested lower value of the AP to the device, eye balled 30 meters to my locations device and changed the setting to 30 saved and rebooted,

I think this fixed my issues above settings and no reboots tested for hours, since then I replaced the MR300 and added MX4300 as the main router and it been running without a reboot for 4 days,

I wonder if the logs I saved about disconnects were part of this issue,
like when a device leaves the area and the AP tries to wait for a call back or ACK acknowledgment
panics when it takes to long and reboots?

But with these settings I think its now resolved, maybe the default for Sensitivity Range / ACK Timing, should not be 500 by default? for QOS ECM SFE setups?

I attached my setup just in case someone else is having similar issue, sorry for the long post,

I will continue to post if anything else happens.

[Alozaros]

I can spot:
bacon interval 103 must be 100, 200, 300 but not 103...
dtim intervals 1 or 2
(i live those both to default) unless i need to change them couse few more routers around...at the same spot

ACK time... the default of 500 must be set 450, 900 (mine), 1350 or 3150 never used 0 for it...

also as the others mentioned QoS doesn't play well with any SFE..

you also never reported back if your power adapter is fine

My WRX36 is rock solid...and runs for days with ECM-NSS + NSS Subsystem enabled...sometimes it has 20+ clients, it also has x4 Vlan's and im using only 2.4Ghz radio on it, my plan is to replace my
edge router R7800, sadly during the flash, SSH CLI update reports 'out of ram but still flashing'(did a few updates before i spotted it)...as sometimes im far away from the unit,CLI update for me is a must...

[foz111]

Never had any random reboots on my wrx36 (r59582), but I don't use the wireless, SFE disabled, NSS Subsystem enabled. Currently 21 days uptime. +1 for checking power supply